-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathetl.html
978 lines (937 loc) · 81.5 KB
/
etl.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
<!DOCTYPE html>
<html lang="" xml:lang="">
<head>
<meta charset="utf-8" />
<meta http-equiv="X-UA-Compatible" content="IE=edge" />
<title>3 ETL | Banco</title>
<meta name="description" content="3 ETL | Banco" />
<meta name="generator" content="bookdown 0.21 and GitBook 2.6.7" />
<meta property="og:title" content="3 ETL | Banco" />
<meta property="og:type" content="book" />
<meta name="twitter:card" content="summary" />
<meta name="twitter:title" content="3 ETL | Banco" />
<meta name="author" content="Carlos, Alberto y Juan Ramón" />
<meta name="viewport" content="width=device-width, initial-scale=1" />
<meta name="apple-mobile-web-app-capable" content="yes" />
<meta name="apple-mobile-web-app-status-bar-style" content="black" />
<link rel="prev" href="análisis-exploratorio-variables.html"/>
<link rel="next" href="modelado.html"/>
<script src="book_assets/jquery-2.2.3/jquery.min.js"></script>
<link href="book_assets/gitbook-2.6.7/css/style.css" rel="stylesheet" />
<link href="book_assets/gitbook-2.6.7/css/plugin-table.css" rel="stylesheet" />
<link href="book_assets/gitbook-2.6.7/css/plugin-bookdown.css" rel="stylesheet" />
<link href="book_assets/gitbook-2.6.7/css/plugin-highlight.css" rel="stylesheet" />
<link href="book_assets/gitbook-2.6.7/css/plugin-search.css" rel="stylesheet" />
<link href="book_assets/gitbook-2.6.7/css/plugin-fontsettings.css" rel="stylesheet" />
<link href="book_assets/gitbook-2.6.7/css/plugin-clipboard.css" rel="stylesheet" />
<script src="book_assets/accessible-code-block-0.0.1/empty-anchor.js"></script>
<link href="book_assets/anchor-sections-1.0/anchor-sections.css" rel="stylesheet" />
<script src="book_assets/anchor-sections-1.0/anchor-sections.js"></script>
<style type="text/css">
code.sourceCode > span { display: inline-block; line-height: 1.25; }
code.sourceCode > span { color: inherit; text-decoration: inherit; }
code.sourceCode > span:empty { height: 1.2em; }
.sourceCode { overflow: visible; }
code.sourceCode { white-space: pre; position: relative; }
pre.sourceCode { margin: 0; }
@media screen {
div.sourceCode { overflow: auto; }
}
@media print {
code.sourceCode { white-space: pre-wrap; }
code.sourceCode > span { text-indent: -5em; padding-left: 5em; }
}
pre.numberSource code
{ counter-reset: source-line 0; }
pre.numberSource code > span
{ position: relative; left: -4em; counter-increment: source-line; }
pre.numberSource code > span > a:first-child::before
{ content: counter(source-line);
position: relative; left: -1em; text-align: right; vertical-align: baseline;
border: none; display: inline-block;
-webkit-touch-callout: none; -webkit-user-select: none;
-khtml-user-select: none; -moz-user-select: none;
-ms-user-select: none; user-select: none;
padding: 0 4px; width: 4em;
color: #aaaaaa;
}
pre.numberSource { margin-left: 3em; border-left: 1px solid #aaaaaa; padding-left: 4px; }
div.sourceCode
{ }
@media screen {
code.sourceCode > span > a:first-child::before { text-decoration: underline; }
}
code span.al { color: #ff0000; font-weight: bold; } /* Alert */
code span.an { color: #60a0b0; font-weight: bold; font-style: italic; } /* Annotation */
code span.at { color: #7d9029; } /* Attribute */
code span.bn { color: #40a070; } /* BaseN */
code span.bu { } /* BuiltIn */
code span.cf { color: #007020; font-weight: bold; } /* ControlFlow */
code span.ch { color: #4070a0; } /* Char */
code span.cn { color: #880000; } /* Constant */
code span.co { color: #60a0b0; font-style: italic; } /* Comment */
code span.cv { color: #60a0b0; font-weight: bold; font-style: italic; } /* CommentVar */
code span.do { color: #ba2121; font-style: italic; } /* Documentation */
code span.dt { color: #902000; } /* DataType */
code span.dv { color: #40a070; } /* DecVal */
code span.er { color: #ff0000; font-weight: bold; } /* Error */
code span.ex { } /* Extension */
code span.fl { color: #40a070; } /* Float */
code span.fu { color: #06287e; } /* Function */
code span.im { } /* Import */
code span.in { color: #60a0b0; font-weight: bold; font-style: italic; } /* Information */
code span.kw { color: #007020; font-weight: bold; } /* Keyword */
code span.op { color: #666666; } /* Operator */
code span.ot { color: #007020; } /* Other */
code span.pp { color: #bc7a00; } /* Preprocessor */
code span.sc { color: #4070a0; } /* SpecialChar */
code span.ss { color: #bb6688; } /* SpecialString */
code span.st { color: #4070a0; } /* String */
code span.va { color: #19177c; } /* Variable */
code span.vs { color: #4070a0; } /* VerbatimString */
code span.wa { color: #60a0b0; font-weight: bold; font-style: italic; } /* Warning */
</style>
</head>
<body>
<div class="book without-animation with-summary font-size-2 font-family-1" data-basepath=".">
<div class="book-summary">
<nav role="navigation">
<ul class="summary">
<li class="chapter" data-level="1" data-path="introducción-y-objetivos.html"><a href="introducción-y-objetivos.html"><i class="fa fa-check"></i><b>1</b> Introducción y objetivos</a><ul>
<li class="chapter" data-level="1.1" data-path="introducción-y-objetivos.html"><a href="introducción-y-objetivos.html#introducción"><i class="fa fa-check"></i><b>1.1</b> Introducción</a></li>
<li class="chapter" data-level="1.2" data-path="introducción-y-objetivos.html"><a href="introducción-y-objetivos.html#objetivos"><i class="fa fa-check"></i><b>1.2</b> Objetivos</a></li>
</ul></li>
<li class="chapter" data-level="2" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html"><i class="fa fa-check"></i><b>2</b> ANÁLISIS EXPLORATORIO VARIABLES</a><ul>
<li class="chapter" data-level="2.1" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#importación-dataset"><i class="fa fa-check"></i><b>2.1</b> Importación dataset</a></li>
<li class="chapter" data-level="2.2" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#división-en-train-y-test"><i class="fa fa-check"></i><b>2.2</b> División en train y test</a></li>
<li class="chapter" data-level="2.3" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#estructura-y-resumen-de-train"><i class="fa fa-check"></i><b>2.3</b> Estructura y resumen de train</a></li>
<li class="chapter" data-level="2.4" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#exploratorio-univariante-de-las-variables-predictoras"><i class="fa fa-check"></i><b>2.4</b> Exploratorio univariante de las variables predictoras</a><ul>
<li class="chapter" data-level="2.4.1" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-age"><i class="fa fa-check"></i><b>2.4.1</b> Variable age</a></li>
<li class="chapter" data-level="2.4.2" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-balance"><i class="fa fa-check"></i><b>2.4.2</b> Variable balance</a></li>
<li class="chapter" data-level="2.4.3" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-job"><i class="fa fa-check"></i><b>2.4.3</b> Variable job</a></li>
<li class="chapter" data-level="2.4.4" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-marital"><i class="fa fa-check"></i><b>2.4.4</b> Variable marital</a></li>
<li class="chapter" data-level="2.4.5" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-education"><i class="fa fa-check"></i><b>2.4.5</b> Variable education</a></li>
<li class="chapter" data-level="2.4.6" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-default"><i class="fa fa-check"></i><b>2.4.6</b> Variable default</a></li>
<li class="chapter" data-level="2.4.7" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-housing"><i class="fa fa-check"></i><b>2.4.7</b> Variable housing</a></li>
<li class="chapter" data-level="2.4.8" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-loan"><i class="fa fa-check"></i><b>2.4.8</b> Variable loan</a></li>
<li class="chapter" data-level="2.4.9" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-contact"><i class="fa fa-check"></i><b>2.4.9</b> Variable contact</a></li>
<li class="chapter" data-level="2.4.10" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-mes"><i class="fa fa-check"></i><b>2.4.10</b> Variable mes</a></li>
<li class="chapter" data-level="2.4.11" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-day"><i class="fa fa-check"></i><b>2.4.11</b> Variable day</a></li>
<li class="chapter" data-level="2.4.12" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-duration"><i class="fa fa-check"></i><b>2.4.12</b> Variable duration</a></li>
<li class="chapter" data-level="2.4.13" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-campaign"><i class="fa fa-check"></i><b>2.4.13</b> Variable campaign</a></li>
<li class="chapter" data-level="2.4.14" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-pdays"><i class="fa fa-check"></i><b>2.4.14</b> Variable pdays</a></li>
<li class="chapter" data-level="2.4.15" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-poutcome"><i class="fa fa-check"></i><b>2.4.15</b> Variable poutcome</a></li>
<li class="chapter" data-level="2.4.16" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-previous"><i class="fa fa-check"></i><b>2.4.16</b> Variable previous</a></li>
<li class="chapter" data-level="2.4.17" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#variable-term_deposit-objetivo"><i class="fa fa-check"></i><b>2.4.17</b> Variable term_deposit (Objetivo)</a></li>
</ul></li>
<li class="chapter" data-level="2.5" data-path="análisis-exploratorio-variables.html"><a href="análisis-exploratorio-variables.html#exploratorio-multivariante"><i class="fa fa-check"></i><b>2.5</b> Exploratorio multivariante</a></li>
</ul></li>
<li class="chapter" data-level="3" data-path="etl.html"><a href="etl.html"><i class="fa fa-check"></i><b>3</b> ETL</a><ul>
<li class="chapter" data-level="3.1" data-path="etl.html"><a href="etl.html#tratamiento-de-datos-faltantes"><i class="fa fa-check"></i><b>3.1</b> Tratamiento de datos faltantes</a></li>
<li class="chapter" data-level="3.2" data-path="etl.html"><a href="etl.html#transformación-de-variables"><i class="fa fa-check"></i><b>3.2</b> Transformación de variables</a><ul>
<li class="chapter" data-level="3.2.1" data-path="etl.html"><a href="etl.html#train-poutcome"><i class="fa fa-check"></i><b>3.2.1</b> Train poutcome</a></li>
<li class="chapter" data-level="3.2.2" data-path="etl.html"><a href="etl.html#variable-previous-1"><i class="fa fa-check"></i><b>3.2.2</b> Variable previous</a></li>
<li class="chapter" data-level="3.2.3" data-path="etl.html"><a href="etl.html#variable-pdays-1"><i class="fa fa-check"></i><b>3.2.3</b> Variable pdays</a></li>
<li class="chapter" data-level="3.2.4" data-path="etl.html"><a href="etl.html#variable-campaign-1"><i class="fa fa-check"></i><b>3.2.4</b> Variable campaign</a></li>
<li class="chapter" data-level="3.2.5" data-path="etl.html"><a href="etl.html#variable-duration-1"><i class="fa fa-check"></i><b>3.2.5</b> Variable duration</a></li>
<li class="chapter" data-level="3.2.6" data-path="etl.html"><a href="etl.html#variable-month"><i class="fa fa-check"></i><b>3.2.6</b> Variable month</a></li>
<li class="chapter" data-level="3.2.7" data-path="etl.html"><a href="etl.html#variable-contact-1"><i class="fa fa-check"></i><b>3.2.7</b> Variable contact</a></li>
<li class="chapter" data-level="3.2.8" data-path="etl.html"><a href="etl.html#variable-balance-1"><i class="fa fa-check"></i><b>3.2.8</b> Variable balance</a></li>
<li class="chapter" data-level="3.2.9" data-path="etl.html"><a href="etl.html#variable-age-1"><i class="fa fa-check"></i><b>3.2.9</b> Variable age</a></li>
</ul></li>
<li class="chapter" data-level="3.3" data-path="etl.html"><a href="etl.html#selección-de-variables"><i class="fa fa-check"></i><b>3.3</b> Selección de variables</a></li>
</ul></li>
<li class="chapter" data-level="4" data-path="modelado.html"><a href="modelado.html"><i class="fa fa-check"></i><b>4</b> Modelado</a></li>
<li class="chapter" data-level="5" data-path="evaluación.html"><a href="evaluación.html"><i class="fa fa-check"></i><b>5</b> Evaluación</a></li>
</ul>
</nav>
</div>
<div class="book-body">
<div class="body-inner">
<div class="book-header" role="navigation">
<h1>
<i class="fa fa-circle-o-notch fa-spin"></i><a href="./">Banco</a>
</h1>
</div>
<div class="page-wrapper" tabindex="-1" role="main">
<div class="page-inner">
<section class="normal" id="section-">
<div id="etl" class="section level1">
<h1><span class="header-section-number">3</span> ETL</h1>
<div id="tratamiento-de-datos-faltantes" class="section level2">
<h2><span class="header-section-number">3.1</span> Tratamiento de datos faltantes</h2>
<p>Las columnas que presentan datos faltantes son <strong>job</strong> y <strong>education</strong> donde venían codificados bajo la categoría unknown. Es lógico pensar que dicha categoría no sean NaNs sino una categoría propia con capacidad informativa por sí sola. Pero en el EDA no hemos observado una distribución diferente a la del resto de categorías, a diferencia de para contact o poutcome donde unknown si parece ser una categoría propia.</p>
<p>El método de imputación utilizado aquí es sencillo. Se utiliza un RandomForest para predecir cada uno de los valores faltantes a partir del resto de variables predictoras que tenemos.</p>
<div class="sourceCode" id="cb378"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb378-1"><a href="etl.html#cb378-1"></a>train =<span class="st"> </span>data[trainIndex ,]</span>
<span id="cb378-2"><a href="etl.html#cb378-2"></a>train[train<span class="op">$</span>education<span class="op">==</span><span class="st">'unknown'</span>,<span class="st">"education"</span>] <-<span class="st"> </span><span class="ot">NA</span></span>
<span id="cb378-3"><a href="etl.html#cb378-3"></a>train[train<span class="op">$</span>job<span class="op">==</span><span class="st">'unknown'</span>,<span class="st">"job"</span>] <-<span class="st"> </span><span class="ot">NA</span></span></code></pre></div>
<div class="sourceCode" id="cb379"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb379-1"><a href="etl.html#cb379-1"></a>train <-<span class="st"> </span>train <span class="op">%>%</span></span>
<span id="cb379-2"><a href="etl.html#cb379-2"></a><span class="st"> </span><span class="kw">mutate</span>(</span>
<span id="cb379-3"><a href="etl.html#cb379-3"></a> <span class="dt">job =</span> <span class="kw">as.factor</span>(job),</span>
<span id="cb379-4"><a href="etl.html#cb379-4"></a> <span class="dt">education =</span> <span class="kw">as.factor</span>(education)</span>
<span id="cb379-5"><a href="etl.html#cb379-5"></a> )</span></code></pre></div>
<div class="sourceCode" id="cb380"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb380-1"><a href="etl.html#cb380-1"></a>d <-<span class="st"> </span><span class="kw">describe</span>(train)</span>
<span id="cb380-2"><a href="etl.html#cb380-2"></a>d[<span class="kw">c</span>(<span class="st">'job'</span>,<span class="st">'education'</span>)]</span></code></pre></div>
<pre><code>## train
##
## 2 Variables 34111 Observations
## --------------------------------------------------------------------------------
## job
## n missing distinct
## 33904 207 11
##
## lowest : admin. blue-collar entrepreneur housemaid management
## highest: self-employed services student technician unemployed
##
## admin. (3892, 0.115), blue-collar (7608, 0.224), entrepreneur (1137, 0.034),
## housemaid (937, 0.028), management (7130, 0.210), retired (1503, 0.044),
## self-employed (1183, 0.035), services (3215, 0.095), student (581, 0.017),
## technician (5763, 0.170), unemployed (955, 0.028)
## --------------------------------------------------------------------------------
## education
## n missing distinct
## 32764 1347 3
##
## Value primary secondary tertiary
## Frequency 5272 17645 9847
## Proportion 0.161 0.539 0.301
## --------------------------------------------------------------------------------</code></pre>
<p>La siguiente visualización nos indica que no parece existir ningún patrón en la falta de valores, el patrón más común es donde a la observación solo le falta el valor para <strong>education</strong>, algo lógico ya que es una variable con el ratio de valores missing más alto.</p>
<div class="sourceCode" id="cb382"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb382-1"><a href="etl.html#cb382-1"></a>aggr_plot <-<span class="st"> </span><span class="kw">aggr</span>(train, <span class="dt">col=</span><span class="kw">c</span>(<span class="st">'navyblue'</span>,<span class="st">'red'</span>), <span class="dt">numbers=</span><span class="ot">TRUE</span>, <span class="dt">sortVars=</span><span class="ot">TRUE</span>,</span>
<span id="cb382-2"><a href="etl.html#cb382-2"></a> <span class="dt">labels=</span><span class="kw">names</span>(train), <span class="dt">cex.axis=</span>.<span class="dv">7</span>, <span class="dt">gap=</span><span class="dv">3</span>, <span class="dt">cex.numbers=</span><span class="fl">0.5</span>,</span>
<span id="cb382-3"><a href="etl.html#cb382-3"></a> <span class="dt">ylab=</span><span class="kw">c</span>(<span class="st">"Histogram of missing data"</span>,<span class="st">"Pattern"</span>))</span></code></pre></div>
<p><img src="bookDown_files/figure-html/unnamed-chunk-145-1.png" width="672" /></p>
<pre><code>##
## Variables sorted by number of missings:
## Variable Count
## education 0.039488728
## job 0.006068424
## age 0.000000000
## marital 0.000000000
## default 0.000000000
## balance 0.000000000
## housing 0.000000000
## loan 0.000000000
## contact 0.000000000
## day 0.000000000
## month 0.000000000
## duration 0.000000000
## campaign 0.000000000
## pdays 0.000000000
## previous 0.000000000
## poutcome 0.000000000
## term_deposit 0.000000000</code></pre>
<p>Realizamos la imputación de valores.</p>
<div class="sourceCode" id="cb384"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb384-1"><a href="etl.html#cb384-1"></a>train_imputed =<span class="st"> </span><span class="kw">missRanger</span>(train, <span class="dt">num.trees =</span> <span class="dv">100</span>)</span></code></pre></div>
<pre><code>##
## Missing value imputation by random forests
##
## Variables to impute: job, education
## Variables used to impute: age, job, marital, education, default, balance, housing, loan, contact, day, month, duration, campaign, pdays, previous, poutcome, term_deposit
## iter 1: ..
## iter 2: ..
## iter 3: ..
## iter 4: ..</code></pre>
<div class="sourceCode" id="cb386"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb386-1"><a href="etl.html#cb386-1"></a><span class="kw">summary</span>(train_imputed)</span></code></pre></div>
<pre><code>## age job marital education
## Min. :18.00 blue-collar:7699 Length:34111 primary : 5505
## 1st Qu.:33.00 management :7199 Class :character secondary:18500
## Median :39.00 technician :5771 Mode :character tertiary :10106
## Mean :40.79 admin. :3895
## 3rd Qu.:48.00 services :3216
## Max. :95.00 retired :1534
## (Other) :4797
## default balance housing loan
## Length:34111 Min. : -8019.0 Length:34111 Length:34111
## Class :character 1st Qu.: 63.5 Class :character Class :character
## Mode :character Median : 432.0 Mode :character Mode :character
## Mean : 1345.1
## 3rd Qu.: 1392.5
## Max. :102127.0
##
## contact day month duration
## Length:34111 Min. : 1.00 Length:34111 Min. : 0.0
## Class :character 1st Qu.: 8.00 Class :character 1st Qu.: 100.0
## Mode :character Median :16.00 Mode :character Median : 176.0
## Mean :15.84 Mean : 255.6
## 3rd Qu.:21.00 3rd Qu.: 315.0
## Max. :31.00 Max. :4918.0
##
## campaign pdays previous poutcome
## Min. : 1.00 Min. : -1.0 Min. : 0.0000 Length:34111
## 1st Qu.: 1.00 1st Qu.: -1.0 1st Qu.: 0.0000 Class :character
## Median : 2.00 Median : -1.0 Median : 0.0000 Mode :character
## Mean : 2.81 Mean : 34.1 Mean : 0.4633
## 3rd Qu.: 3.00 3rd Qu.: -1.0 3rd Qu.: 0.0000
## Max. :63.00 Max. :536.0 Max. :275.0000
##
## term_deposit
## Length:34111
## Class :character
## Mode :character
##
##
##
## </code></pre>
<div class="sourceCode" id="cb388"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb388-1"><a href="etl.html#cb388-1"></a><span class="kw">describe</span>(train_imputed)</span></code></pre></div>
<pre><code>## train_imputed
##
## 17 Variables 34111 Observations
## --------------------------------------------------------------------------------
## age
## n missing distinct Info Mean Gmd .05 .10
## 34111 0 74 0.999 40.79 11.47 27 29
## .25 .50 .75 .90 .95
## 33 39 48 56 58
##
## lowest : 18 19 20 21 22, highest: 87 89 90 92 95
## --------------------------------------------------------------------------------
## job
## n missing distinct
## 34111 0 11
##
## lowest : admin. blue-collar entrepreneur housemaid management
## highest: self-employed services student technician unemployed
##
## admin. (3895, 0.114), blue-collar (7699, 0.226), entrepreneur (1137, 0.033),
## housemaid (939, 0.028), management (7199, 0.211), retired (1534, 0.045),
## self-employed (1183, 0.035), services (3216, 0.094), student (582, 0.017),
## technician (5771, 0.169), unemployed (956, 0.028)
## --------------------------------------------------------------------------------
## marital
## n missing distinct
## 34111 0 3
##
## Value divorced married single
## Frequency 3982 20680 9449
## Proportion 0.117 0.606 0.277
## --------------------------------------------------------------------------------
## education
## n missing distinct
## 34111 0 3
##
## Value primary secondary tertiary
## Frequency 5505 18500 10106
## Proportion 0.161 0.542 0.296
## --------------------------------------------------------------------------------
## default
## n missing distinct
## 34111 0 2
##
## Value no yes
## Frequency 33463 648
## Proportion 0.981 0.019
## --------------------------------------------------------------------------------
## balance
## n missing distinct Info Mean Gmd .05 .10
## 34111 0 6531 0.999 1345 2045 -191.0 0.0
## .25 .50 .75 .90 .95
## 63.5 432.0 1392.5 3570.0 5771.0
##
## lowest : -8019 -6847 -4057 -3372 -3313, highest: 64343 66721 71188 81204 102127
## --------------------------------------------------------------------------------
## housing
## n missing distinct
## 34111 0 2
##
## Value no yes
## Frequency 14509 19602
## Proportion 0.425 0.575
## --------------------------------------------------------------------------------
## loan
## n missing distinct
## 34111 0 2
##
## Value no yes
## Frequency 28463 5648
## Proportion 0.834 0.166
## --------------------------------------------------------------------------------
## contact
## n missing distinct
## 34111 0 3
##
## Value cellular telephone unknown
## Frequency 21769 2097 10245
## Proportion 0.638 0.061 0.300
## --------------------------------------------------------------------------------
## day
## n missing distinct Info Mean Gmd .05 .10
## 34111 0 31 0.998 15.84 9.544 3 5
## .25 .50 .75 .90 .95
## 8 16 21 28 30
##
## lowest : 1 2 3 4 5, highest: 27 28 29 30 31
## --------------------------------------------------------------------------------
## month
## n missing distinct
## 34111 0 12
##
## lowest : apr aug dec feb jan, highest: mar may nov oct sep
##
## Value apr aug dec feb jan jul jun mar may nov oct
## Frequency 2142 4813 171 1823 969 5244 4125 214 10852 3113 419
## Proportion 0.063 0.141 0.005 0.053 0.028 0.154 0.121 0.006 0.318 0.091 0.012
##
## Value sep
## Frequency 226
## Proportion 0.007
## --------------------------------------------------------------------------------
## duration
## n missing distinct Info Mean Gmd .05 .10
## 34111 0 1485 1 255.6 235 35 58
## .25 .50 .75 .90 .95
## 100 176 315 544 749
##
## lowest : 0 2 3 4 5, highest: 3284 3322 3366 3881 4918
## --------------------------------------------------------------------------------
## campaign
## n missing distinct Info Mean Gmd .05 .10
## 34111 0 45 0.921 2.81 2.435 1 1
## .25 .50 .75 .90 .95
## 1 2 3 5 8
##
## lowest : 1 2 3 4 5, highest: 43 44 55 58 63
## --------------------------------------------------------------------------------
## pdays
## n missing distinct Info Mean Gmd .05 .10
## 34111 0 409 0.393 34.1 62.25 -1 -1
## .25 .50 .75 .90 .95
## -1 -1 -1 175 301
##
## lowest : -1 1 2 4 5, highest: 495 515 518 520 536
## --------------------------------------------------------------------------------
## previous
## n missing distinct Info Mean Gmd .05 .10
## 34111 0 36 0.393 0.4633 0.8488 0 0
## .25 .50 .75 .90 .95
## 0 0 0 1 3
##
## lowest : 0 1 2 3 4, highest: 37 38 51 58 275
## --------------------------------------------------------------------------------
## poutcome
## n missing distinct
## 34111 0 4
##
## Value failure other success unknown
## Frequency 3401 1201 625 28884
## Proportion 0.100 0.035 0.018 0.847
## --------------------------------------------------------------------------------
## term_deposit
## n missing distinct
## 34111 0 2
##
## Value no yes
## Frequency 30938 3173
## Proportion 0.907 0.093
## --------------------------------------------------------------------------------</code></pre>
<div class="sourceCode" id="cb390"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb390-1"><a href="etl.html#cb390-1"></a>train =<span class="st"> </span>train_imputed</span>
<span id="cb390-2"><a href="etl.html#cb390-2"></a>train<span class="op">$</span>term_deposit =<span class="st"> </span><span class="kw">factor</span>(train<span class="op">$</span>term_deposit)</span></code></pre></div>
</div>
<div id="transformación-de-variables" class="section level2">
<h2><span class="header-section-number">3.2</span> Transformación de variables</h2>
<p>Una vez se han obtenido las imputaciones de los missing values vamos a proceder a realizar una selección de variables. El procedimiento es el siguiente, nos enfocamos en la variable con menor importancia, entendiendo que una variable es importante si el peso dado en la regresión se encuentra alejado de 0. A esta variable tratamos de buscarla un tratamiento diferente que pueda mejorar el desempeño del modelo así como aumentar su importancia dentro del ranking.</p>
<p>Consideramos que una transformación mejora el desempeño si el F-score calculado mediante 10-fold cross validation mejora. Esta validación cruzada se ha aplicado en el conjunto de train, nunca en test para evitar sobreajustes.</p>
<div class="sourceCode" id="cb391"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb391-1"><a href="etl.html#cb391-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb392"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb392-1"><a href="etl.html#cb392-1"></a><span class="co"># Train the model</span></span>
<span id="cb392-2"><a href="etl.html#cb392-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> train, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb392-3"><a href="etl.html#cb392-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb394"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb394-1"><a href="etl.html#cb394-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb394-2"><a href="etl.html#cb394-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34111 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30701, 30699, 30700, 30700, 30699, 30699, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9889767 0.9320116 0.981382 0.9560568</code></pre>
<div class="sourceCode" id="cb396"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb396-1"><a href="etl.html#cb396-1"></a>importance <-<span class="st"> </span><span class="kw">varImp</span>(model, <span class="dt">scale=</span><span class="ot">FALSE</span>)</span>
<span id="cb396-2"><a href="etl.html#cb396-2"></a><span class="co"># summarize importance</span></span>
<span id="cb396-3"><a href="etl.html#cb396-3"></a><span class="kw">print</span>(<span class="kw">rownames</span>(importance<span class="op">$</span>importance))</span></code></pre></div>
<pre><code>## [1] "age" "`jobblue-collar`" "jobentrepreneur"
## [4] "jobhousemaid" "jobmanagement" "jobretired"
## [7] "`jobself-employed`" "jobservices" "jobstudent"
## [10] "jobtechnician" "jobunemployed" "maritalmarried"
## [13] "maritalsingle" "educationsecondary" "educationtertiary"
## [16] "defaultyes" "balance" "housingyes"
## [19] "loanyes" "contacttelephone" "contactunknown"
## [22] "day" "monthaug" "monthdec"
## [25] "monthfeb" "monthjan" "monthjul"
## [28] "monthjun" "monthmar" "monthmay"
## [31] "monthnov" "monthoct" "monthsep"
## [34] "duration" "campaign" "pdays"
## [37] "previous" "poutcomeother" "poutcomesuccess"
## [40] "poutcomeunknown"</code></pre>
<div id="train-poutcome" class="section level3">
<h3><span class="header-section-number">3.2.1</span> Train poutcome</h3>
<p>La variable categórica que informa del resultado en la última campaña es la menos informativa, vamos a probar a realizar una re-categorización donde agrupemos en una categoría las que tienen una menor probabilidad de contratar (failure, other y unknown) y en otra categoría success.</p>
<div class="sourceCode" id="cb398"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb398-1"><a href="etl.html#cb398-1"></a>poutcome_aprox <-<span class="st"> </span>train</span>
<span id="cb398-2"><a href="etl.html#cb398-2"></a>poutcome_aprox[poutcome_aprox<span class="op">$</span>poutcome<span class="op">==</span><span class="st">"failure"</span> <span class="op">|</span><span class="st"> </span>poutcome_aprox<span class="op">$</span>poutcome<span class="op">==</span><span class="st">"other"</span> <span class="op">|</span><span class="st"> </span>poutcome_aprox<span class="op">$</span>poutcome<span class="op">==</span><span class="st">"unknown"</span>,<span class="st">"poutcome"</span>] <-<span class="st"> "bad_poutcome"</span></span>
<span id="cb398-3"><a href="etl.html#cb398-3"></a>poutcome_aprox[poutcome_aprox<span class="op">$</span>poutcome<span class="op">==</span><span class="st">"success"</span>,<span class="st">"poutcome"</span>] <-<span class="st"> "success_poutcome"</span></span></code></pre></div>
<div class="sourceCode" id="cb399"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb399-1"><a href="etl.html#cb399-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb400"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb400-1"><a href="etl.html#cb400-1"></a><span class="co"># Train the model</span></span>
<span id="cb400-2"><a href="etl.html#cb400-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> poutcome_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb400-3"><a href="etl.html#cb400-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb402"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb402-1"><a href="etl.html#cb402-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb402-2"><a href="etl.html#cb402-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34111 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30699, 30701, 30700, 30700, 30700, 30700, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9888656 0.9318751 0.9814791 0.9560322</code></pre>
<p>Con esta nueva distribución de la variable no parece haber mejora ni cobrar mayor importancia en el ranking de variables, por lo que el siguiente paso sería a probar a eliminar dicha variable.</p>
</div>
<div id="variable-previous-1" class="section level3">
<h3><span class="header-section-number">3.2.2</span> Variable previous</h3>
<p>La siguiente variable con menor importancia es previous, variable cuantitativa que muestra el número de contactos que realizó el cliente en campañas anteriores. En este caso, tal y como hicimos en el exploratorio, vamos a categorizar esta variable en dos categorías: Sin contacto previo y contacto previo.</p>
<div class="sourceCode" id="cb404"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb404-1"><a href="etl.html#cb404-1"></a>previous_aprox <-<span class="st"> </span>train <span class="op">%>%</span><span class="st"> </span><span class="kw">mutate</span>(<span class="dt">previous_categorica =</span> <span class="kw">cut</span>(previous, <span class="dt">breaks =</span> <span class="kw">c</span>(<span class="dv">0</span>, <span class="dv">1</span>, <span class="dv">300</span>), <span class="dt">right =</span> <span class="ot">FALSE</span>, <span class="dt">include.lowest =</span> <span class="ot">TRUE</span>, <span class="dt">labels =</span> <span class="kw">c</span>(<span class="st">'Sin Contacto previo'</span>, <span class="st">'Contacto previo'</span>)))</span>
<span id="cb404-2"><a href="etl.html#cb404-2"></a>previous_aprox<span class="op">$</span>previous <-<span class="st"> </span><span class="ot">NULL</span></span></code></pre></div>
<div class="sourceCode" id="cb405"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb405-1"><a href="etl.html#cb405-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb406"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb406-1"><a href="etl.html#cb406-1"></a><span class="co"># Train the model</span></span>
<span id="cb406-2"><a href="etl.html#cb406-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> previous_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb406-3"><a href="etl.html#cb406-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb408"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb408-1"><a href="etl.html#cb408-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb408-2"><a href="etl.html#cb408-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34111 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30700, 30700, 30699, 30700, 30700, 30699, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9889662 0.9319703 0.9811883 0.9559429</code></pre>
<p>Esta transformación de la variable previous no parece mejorar el desempeño del modelo.</p>
</div>
<div id="variable-pdays-1" class="section level3">
<h3><span class="header-section-number">3.2.3</span> Variable pdays</h3>
<p>La siguiente variable en el ranking es pdays, variable cuantitativa que muestra los días que han pasado desde que se contactó con él en la anterior campaña. En el exploratorio se ha propuesto ya una categorización que vamos a comprobar ahora.</p>
<div class="sourceCode" id="cb410"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb410-1"><a href="etl.html#cb410-1"></a>pdays_aprox =<span class="st"> </span>train <span class="op">%>%</span><span class="st"> </span><span class="kw">mutate</span>(<span class="dt">pdays_categorica =</span> <span class="kw">cut</span>(pdays, <span class="dt">breaks =</span> <span class="kw">c</span>(<span class="op">-</span><span class="dv">1</span>, <span class="dv">0</span>, <span class="dv">210</span>, <span class="dv">536</span>), <span class="dt">right =</span> <span class="ot">FALSE</span>, <span class="dt">include.lowest =</span> <span class="ot">TRUE</span>, <span class="dt">labels =</span> <span class="kw">c</span>(<span class="st">'No contacto'</span>, <span class="st">'<7 Meses'</span>, <span class="st">'>7 Meses'</span>)))</span>
<span id="cb410-2"><a href="etl.html#cb410-2"></a>pdays_aprox<span class="op">$</span>pdays <-<span class="st"> </span><span class="ot">NULL</span></span></code></pre></div>
<div class="sourceCode" id="cb411"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb411-1"><a href="etl.html#cb411-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb412"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb412-1"><a href="etl.html#cb412-1"></a><span class="co"># Train the model</span></span>
<span id="cb412-2"><a href="etl.html#cb412-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> pdays_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb412-3"><a href="etl.html#cb412-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb414"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb414-1"><a href="etl.html#cb414-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb414-2"><a href="etl.html#cb414-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34111 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30701, 30700, 30700, 30699, 30700, 30700, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9889236 0.9319423 0.9811556 0.9559117</code></pre>
<p>Con la categorización de la variable pdays el modelo no parece mejorar.</p>
</div>
<div id="variable-campaign-1" class="section level3">
<h3><span class="header-section-number">3.2.4</span> Variable campaign</h3>
<p>La siguiente variable es campaign, variable cuantitativa que muestra el número de contactos en esta campaña. Ya en el EDA se ha propuesto una categorización que vamos a analizar a continuación.</p>
<div class="sourceCode" id="cb416"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb416-1"><a href="etl.html#cb416-1"></a>campaign_aprox =<span class="st"> </span>train <span class="op">%>%</span><span class="st"> </span><span class="kw">mutate</span>(<span class="dt">campaign_binaria =</span> <span class="kw">cut</span>(campaign, <span class="dt">breaks =</span> <span class="kw">c</span>(<span class="dv">1</span>, <span class="dv">2</span>, <span class="dv">300</span>), <span class="dt">right =</span> <span class="ot">FALSE</span>, <span class="dt">include.lowest =</span> <span class="ot">TRUE</span>, <span class="dt">labels =</span> <span class="kw">c</span>(<span class="st">'1Contacto'</span>, <span class="st">'+1Contacto'</span>)))</span>
<span id="cb416-2"><a href="etl.html#cb416-2"></a>campaign_aprox<span class="op">$</span>campaign <-<span class="st"> </span><span class="ot">NULL</span></span></code></pre></div>
<div class="sourceCode" id="cb417"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb417-1"><a href="etl.html#cb417-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb418"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb418-1"><a href="etl.html#cb418-1"></a><span class="co"># Train the model</span></span>
<span id="cb418-2"><a href="etl.html#cb418-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> campaign_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb418-3"><a href="etl.html#cb418-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb420"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb420-1"><a href="etl.html#cb420-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb420-2"><a href="etl.html#cb420-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34111 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30700, 30701, 30699, 30700, 30700, 30700, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9889353 0.9322811 0.9815437 0.956275</code></pre>
<p>El nuevo tratamiento de la variable campaign mejora el desempeño del modelo.</p>
<div class="sourceCode" id="cb422"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb422-1"><a href="etl.html#cb422-1"></a>train <-<span class="st"> </span>campaign_aprox</span></code></pre></div>
</div>
<div id="variable-duration-1" class="section level3">
<h3><span class="header-section-number">3.2.5</span> Variable duration</h3>
<p>La siguiente variable es duration, variable cuantitativa que muestra la duración en segundos del último contacto mantenido con el cliente. En el EDA se ha propuesto realizar una transformación logarítmica a la variable para tratar de mejorar las diferencias entre la distribución de esta variable en los casos de contratación de los que no.</p>
<div class="sourceCode" id="cb423"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb423-1"><a href="etl.html#cb423-1"></a>duration_aprox <-<span class="st"> </span>train[train<span class="op">$</span>duration<span class="op">></span><span class="dv">0</span>,]</span>
<span id="cb423-2"><a href="etl.html#cb423-2"></a>duration_aprox<span class="op">$</span>duration <-<span class="st"> </span><span class="kw">log10</span>(duration_aprox<span class="op">$</span>duration)</span></code></pre></div>
<div class="sourceCode" id="cb424"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb424-1"><a href="etl.html#cb424-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb425"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb425-1"><a href="etl.html#cb425-1"></a><span class="co"># Train the model</span></span>
<span id="cb425-2"><a href="etl.html#cb425-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> duration_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb425-3"><a href="etl.html#cb425-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb427"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb427-1"><a href="etl.html#cb427-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb427-2"><a href="etl.html#cb427-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34108 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30697, 30697, 30698, 30698, 30698, 30696, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9887624 0.932604 0.9812509 0.9563054</code></pre>
<p>Este tratamiento parece mejorar el desempeño del modelo.</p>
<div class="sourceCode" id="cb429"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb429-1"><a href="etl.html#cb429-1"></a>train <-<span class="st"> </span>duration_aprox</span></code></pre></div>
</div>
<div id="variable-month" class="section level3">
<h3><span class="header-section-number">3.2.6</span> Variable month</h3>
<p>Variable categórica que indica el mes en el que se produce el contacto actual con el cliente. Vamos a re-categorizar la variable en dos categorias en función de si es un mes con alto índice de contrataciones o no.</p>
<div class="sourceCode" id="cb430"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb430-1"><a href="etl.html#cb430-1"></a>month_aprox <-<span class="st"> </span>train</span></code></pre></div>
<div class="sourceCode" id="cb431"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb431-1"><a href="etl.html#cb431-1"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"jan"</span>,<span class="st">"month"</span>] <-<span class="st"> "no_month"</span></span>
<span id="cb431-2"><a href="etl.html#cb431-2"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"feb"</span>,<span class="st">"month"</span>] <-<span class="st"> "no_month"</span></span>
<span id="cb431-3"><a href="etl.html#cb431-3"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"mar"</span>,<span class="st">"month"</span>] <-<span class="st"> "yes_month"</span></span>
<span id="cb431-4"><a href="etl.html#cb431-4"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"apr"</span>,<span class="st">"month"</span>] <-<span class="st"> "no_month"</span></span>
<span id="cb431-5"><a href="etl.html#cb431-5"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"may"</span>,<span class="st">"month"</span>] <-<span class="st"> "no_month"</span></span>
<span id="cb431-6"><a href="etl.html#cb431-6"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"jun"</span>,<span class="st">"month"</span>] <-<span class="st"> "no_month"</span></span>
<span id="cb431-7"><a href="etl.html#cb431-7"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"jul"</span>,<span class="st">"month"</span>] <-<span class="st"> "no_month"</span></span>
<span id="cb431-8"><a href="etl.html#cb431-8"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"aug"</span>,<span class="st">"month"</span>] <-<span class="st"> "no_month"</span></span>
<span id="cb431-9"><a href="etl.html#cb431-9"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"sep"</span>,<span class="st">"month"</span>] <-<span class="st"> "yes_month"</span></span>
<span id="cb431-10"><a href="etl.html#cb431-10"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"oct"</span>,<span class="st">"month"</span>] <-<span class="st"> "yes_month"</span></span>
<span id="cb431-11"><a href="etl.html#cb431-11"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"nov"</span>,<span class="st">"month"</span>] <-<span class="st"> "no_month"</span></span>
<span id="cb431-12"><a href="etl.html#cb431-12"></a>month_aprox[month_aprox<span class="op">$</span>month<span class="op">==</span><span class="st">"dec"</span>,<span class="st">"month"</span>] <-<span class="st"> "yes_month"</span></span></code></pre></div>
<div class="sourceCode" id="cb432"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb432-1"><a href="etl.html#cb432-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb433"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb433-1"><a href="etl.html#cb433-1"></a><span class="co"># Train the model</span></span>
<span id="cb433-2"><a href="etl.html#cb433-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> month_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb433-3"><a href="etl.html#cb433-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb435"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb435-1"><a href="etl.html#cb435-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb435-2"><a href="etl.html#cb435-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34108 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30696, 30698, 30698, 30697, 30696, 30697, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9878167 0.9303059 0.9820266 0.9554642</code></pre>
<p>Este nuevo tratamiento no mejora el desempeño del modelo.</p>
</div>
<div id="variable-contact-1" class="section level3">
<h3><span class="header-section-number">3.2.7</span> Variable contact</h3>
<p>Variable categórica que indica la forma en la que se contactó con el cliente. Vamos a tratar de reducir las categorías a escala binaria en función de si es unknown o no.</p>
<div class="sourceCode" id="cb437"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb437-1"><a href="etl.html#cb437-1"></a>contact_aprox <-<span class="st"> </span>train</span>
<span id="cb437-2"><a href="etl.html#cb437-2"></a>contact_aprox[contact_aprox<span class="op">$</span>contact<span class="op">==</span><span class="st">"unknown"</span>,<span class="st">"contact"</span>] <-<span class="st"> "unknown"</span></span>
<span id="cb437-3"><a href="etl.html#cb437-3"></a>contact_aprox[contact_aprox<span class="op">$</span>contact<span class="op">!=</span><span class="st">"unknown"</span>,<span class="st">"contact"</span>] <-<span class="st"> "not_unknown"</span></span></code></pre></div>
<div class="sourceCode" id="cb438"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb438-1"><a href="etl.html#cb438-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb439"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb439-1"><a href="etl.html#cb439-1"></a><span class="co"># Train the model</span></span>
<span id="cb439-2"><a href="etl.html#cb439-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> contact_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb439-3"><a href="etl.html#cb439-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb441"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb441-1"><a href="etl.html#cb441-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb441-2"><a href="etl.html#cb441-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34108 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30696, 30698, 30697, 30696, 30698, 30698, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9887404 0.9327847 0.9810894 0.9563258</code></pre>
<p>Este nuevo tratamiento parece mejorar el desempeño del modelo.</p>
<div class="sourceCode" id="cb443"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb443-1"><a href="etl.html#cb443-1"></a>train <-<span class="st"> </span>contact_aprox</span></code></pre></div>
</div>
<div id="variable-balance-1" class="section level3">
<h3><span class="header-section-number">3.2.8</span> Variable balance</h3>
<p>Variable cuantitativa que indica el saldo del cliente. En el exploratorio se ha propuesto categorizar esta variable en 4 categorías (saldo negativo, bajo, medio y alto).</p>
<div class="sourceCode" id="cb444"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb444-1"><a href="etl.html#cb444-1"></a>balance_aprox =<span class="st"> </span>train <span class="op">%>%</span><span class="st"> </span><span class="kw">mutate</span>(<span class="dt">balance_categorica =</span> <span class="kw">cut</span>(balance, <span class="dt">breaks =</span> <span class="kw">c</span>(<span class="op">-</span><span class="dv">10000</span>, <span class="dv">0</span>, <span class="dv">2000</span>, <span class="dv">10000</span>, <span class="dv">105000</span>), <span class="dt">right =</span> <span class="ot">FALSE</span>, <span class="dt">labels =</span> <span class="kw">c</span>(<span class="st">'Negativo'</span>,<span class="st">'Bajo'</span>,<span class="st">'Medio'</span>, <span class="st">'Alto'</span>)))</span></code></pre></div>
<div class="sourceCode" id="cb445"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb445-1"><a href="etl.html#cb445-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb446"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb446-1"><a href="etl.html#cb446-1"></a><span class="co"># Train the model</span></span>
<span id="cb446-2"><a href="etl.html#cb446-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> balance_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb446-3"><a href="etl.html#cb446-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb448"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb448-1"><a href="etl.html#cb448-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb448-2"><a href="etl.html#cb448-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34108 samples
## 17 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30697, 30697, 30698, 30698, 30696, 30698, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.9888094 0.9327786 0.9809278 0.9562439</code></pre>
<p>Este tratamiento no aporta beneficio al modelo.</p>
</div>
<div id="variable-age-1" class="section level3">
<h3><span class="header-section-number">3.2.9</span> Variable age</h3>
<p>Variable cuantitativa que indica la edad del cliente. En el exploratorio se ha propuesto categorizar esta variable en 4 grupos (ancianos, mayores, joven y mediana edad).</p>
<div class="sourceCode" id="cb450"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb450-1"><a href="etl.html#cb450-1"></a>age_aprox =<span class="st"> </span>train <span class="op">%>%</span><span class="st"> </span><span class="kw">mutate</span>(<span class="dt">age_categorica =</span> <span class="kw">cut</span>(age, <span class="dt">breaks =</span> <span class="kw">c</span>(<span class="dv">0</span>, <span class="dv">29</span>, <span class="dv">44</span>, <span class="dv">59</span>, <span class="dv">100</span>), <span class="dt">right =</span> <span class="ot">TRUE</span>, <span class="dt">labels =</span> <span class="kw">c</span>(<span class="st">'Joven'</span>,<span class="st">'MedianaEdad'</span>,<span class="st">'Mayores'</span>,<span class="st">'Ancianos'</span>)))</span>
<span id="cb450-2"><a href="etl.html#cb450-2"></a>age_aprox<span class="op">$</span>age <-<span class="st"> </span><span class="ot">NULL</span></span></code></pre></div>
<div class="sourceCode" id="cb451"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb451-1"><a href="etl.html#cb451-1"></a>train.control <-<span class="st"> </span><span class="kw">trainControl</span>(<span class="dt">method =</span> <span class="st">"cv"</span>, <span class="dt">number =</span> <span class="dv">10</span>, <span class="dt">summaryFunction=</span>prSummary,<span class="dt">classProbs =</span> <span class="ot">TRUE</span>)</span></code></pre></div>
<div class="sourceCode" id="cb452"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb452-1"><a href="etl.html#cb452-1"></a><span class="co"># Train the model</span></span>
<span id="cb452-2"><a href="etl.html#cb452-2"></a>model <-<span class="st"> </span><span class="kw">train</span>(term_deposit <span class="op">~</span>., <span class="dt">data =</span> age_aprox, <span class="dt">method =</span> <span class="st">"glm"</span>,</span>
<span id="cb452-3"><a href="etl.html#cb452-3"></a> <span class="dt">trControl =</span> train.control)</span></code></pre></div>
<pre><code>## Warning in train.default(x, y, weights = w, ...): The metric "Accuracy" was not
## in the result set. AUC will be used instead.</code></pre>
<div class="sourceCode" id="cb454"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb454-1"><a href="etl.html#cb454-1"></a><span class="co"># Summarize the results</span></span>
<span id="cb454-2"><a href="etl.html#cb454-2"></a><span class="kw">print</span>(model)</span></code></pre></div>
<pre><code>## Generalized Linear Model
##
## 34108 samples
## 16 predictor
## 2 classes: 'no', 'yes'
##
## No pre-processing
## Resampling: Cross-Validated (10 fold)
## Summary of sample sizes: 30697, 30697, 30696, 30698, 30697, 30698, ...
## Resampling results:
##
## AUC Precision Recall F
## 0.989032 0.9332993 0.9809925 0.9565504</code></pre>
<p>En este caso el modelo mejora con este tratamiento.</p>
<div class="sourceCode" id="cb456"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb456-1"><a href="etl.html#cb456-1"></a>train <-<span class="st"> </span>age_aprox</span></code></pre></div>
</div>
</div>
<div id="selección-de-variables" class="section level2">
<h2><span class="header-section-number">3.3</span> Selección de variables</h2>
<p>Una vez hemos decidido las transformaciones de variables a aplicar vamos a seleccionar las variables predictoras que utilizar en el modelo. Para ello se han utilizado tres métodos de selección, los tres de tipo <strong>wrapper</strong>. Estas aproximaciones son bakward, forward y both. Las tres aproximaciones se basan en una búsqueda de la mejor configuración de variables en base a optimizar el AIC del modelo.</p>
<div class="sourceCode" id="cb457"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb457-1"><a href="etl.html#cb457-1"></a>model <-<span class="st"> </span><span class="kw">glm</span>(term_deposit <span class="op">~</span><span class="st"> </span>., <span class="dt">data =</span> train,<span class="dt">family=</span><span class="st">'binomial'</span>)</span></code></pre></div>
<div class="sourceCode" id="cb458"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb458-1"><a href="etl.html#cb458-1"></a>backward_selection <-<span class="st"> </span><span class="kw">stepAIC</span>(model, <span class="dt">trace=</span><span class="ot">FALSE</span>, <span class="dt">direction=</span><span class="st">"backward"</span>)</span></code></pre></div>
<div class="sourceCode" id="cb459"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb459-1"><a href="etl.html#cb459-1"></a><span class="kw">summary</span>(backward_selection)</span></code></pre></div>
<pre><code>##
## Call:
## glm(formula = term_deposit ~ job + marital + education + balance +
## housing + loan + contact + day + month + duration + pdays +
## poutcome + campaign_binaria + age_categorica, family = "binomial",
## data = train)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -3.3087 -0.3346 -0.1702 -0.0797 4.3082
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.225e+01 2.930e-01 -41.802 < 2e-16 ***
## jobblue-collar -2.584e-01 8.881e-02 -2.910 0.003617 **
## jobentrepreneur -5.423e-01 1.572e-01 -3.450 0.000562 ***
## jobhousemaid -7.257e-01 1.763e-01 -4.115 3.87e-05 ***
## jobmanagement -2.481e-01 9.433e-02 -2.630 0.008529 **
## jobretired -2.328e-01 1.347e-01 -1.728 0.083902 .
## jobself-employed -4.134e-01 1.411e-01 -2.931 0.003380 **
## jobservices -2.470e-01 1.034e-01 -2.389 0.016897 *
## jobstudent 2.538e-01 1.527e-01 1.662 0.096551 .
## jobtechnician -1.785e-01 8.748e-02 -2.040 0.041355 *
## jobunemployed -3.030e-01 1.448e-01 -2.092 0.036432 *
## maritalmarried -2.781e-01 7.185e-02 -3.871 0.000109 ***
## maritalsingle -1.306e-01 8.176e-02 -1.597 0.110198
## educationsecondary 4.173e-02 7.754e-02 0.538 0.590411
## educationtertiary 3.712e-01 9.259e-02 4.009 6.10e-05 ***
## balance 1.759e-05 6.565e-06 2.679 0.007377 **
## housingyes -5.076e-01 5.540e-02 -9.161 < 2e-16 ***
## loanyes -2.630e-01 7.030e-02 -3.741 0.000183 ***
## contactunknown -1.669e+00 8.637e-02 -19.323 < 2e-16 ***
## day 1.280e-02 3.273e-03 3.912 9.17e-05 ***
## monthaug -4.871e-01 1.018e-01 -4.786 1.70e-06 ***
## monthdec 8.613e-01 2.130e-01 4.044 5.25e-05 ***
## monthfeb -7.919e-02 1.247e-01 -0.635 0.525266
## monthjan -1.699e+00 1.814e-01 -9.363 < 2e-16 ***
## monthjul -8.772e-01 9.904e-02 -8.857 < 2e-16 ***
## monthjun 8.273e-01 1.214e-01 6.815 9.45e-12 ***
## monthmar 2.741e+00 1.847e-01 14.835 < 2e-16 ***
## monthmay -1.108e-01 9.160e-02 -1.209 0.226588
## monthnov -6.073e-01 1.039e-01 -5.846 5.04e-09 ***
## monthoct 1.671e+00 1.498e-01 11.156 < 2e-16 ***
## monthsep 1.190e+00 1.933e-01 6.157 7.41e-10 ***
## duration 4.826e+00 8.355e-02 57.764 < 2e-16 ***
## pdays -3.077e-03 5.098e-04 -6.036 1.58e-09 ***
## poutcomeother 6.015e-02 1.293e-01 0.465 0.641801
## poutcomesuccess 1.837e+00 1.284e-01 14.307 < 2e-16 ***
## poutcomeunknown -7.150e-01 1.354e-01 -5.281 1.28e-07 ***
## campaign_binaria+1Contacto -3.019e-01 4.784e-02 -6.310 2.79e-10 ***
## age_categoricaMedianaEdad -3.433e-01 7.357e-02 -4.666 3.08e-06 ***
## age_categoricaMayores -3.929e-01 8.663e-02 -4.535 5.76e-06 ***
## age_categoricaAncianos 4.974e-01 1.455e-01 3.419 0.000629 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 21112 on 34107 degrees of freedom
## Residual deviance: 13348 on 34068 degrees of freedom
## AIC: 13428
##
## Number of Fisher Scoring iterations: 7</code></pre>
<div class="sourceCode" id="cb461"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb461-1"><a href="etl.html#cb461-1"></a>empty_model <-<span class="st"> </span><span class="kw">glm</span>(term_deposit <span class="op">~</span><span class="st"> </span><span class="dv">1</span>, <span class="dt">data=</span>train, <span class="dt">family=</span><span class="st">'binomial'</span>)</span>
<span id="cb461-2"><a href="etl.html#cb461-2"></a>horizonte <-<span class="st"> </span><span class="kw">formula</span>(term_deposit <span class="op">~</span><span class="st"> </span>age_categorica <span class="op">+</span><span class="st"> </span>job <span class="op">+</span><span class="st"> </span>marital <span class="op">+</span><span class="st"> </span>education <span class="op">+</span><span class="st"> </span>default <span class="op">+</span><span class="st"> </span>balance <span class="op">+</span><span class="st"> </span>housing <span class="op">+</span><span class="st"> </span>loan <span class="op">+</span><span class="st"> </span>contact <span class="op">+</span><span class="st"> </span>day <span class="op">+</span><span class="st"> </span>month <span class="op">+</span><span class="st"> </span>duration <span class="op">+</span><span class="st"> </span>campaign_binaria <span class="op">+</span><span class="st"> </span>pdays <span class="op">+</span><span class="st"> </span>previous <span class="op">+</span><span class="st"> </span>poutcome)</span></code></pre></div>
<div class="sourceCode" id="cb462"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb462-1"><a href="etl.html#cb462-1"></a>forward_selection <-<span class="st"> </span><span class="kw">stepAIC</span>(empty_model, <span class="dt">trace=</span><span class="ot">FALSE</span>, <span class="dt">direction=</span><span class="st">"forward"</span>, <span class="dt">scope=</span>horizonte)</span></code></pre></div>
<div class="sourceCode" id="cb463"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb463-1"><a href="etl.html#cb463-1"></a><span class="kw">summary</span>(forward_selection)</span></code></pre></div>
<pre><code>##
## Call:
## glm(formula = term_deposit ~ duration + month + contact + poutcome +
## housing + age_categorica + education + pdays + campaign_binaria +
## job + loan + marital + day + balance, family = "binomial",
## data = train)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -3.3087 -0.3346 -0.1702 -0.0797 4.3082
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.225e+01 2.930e-01 -41.802 < 2e-16 ***
## duration 4.826e+00 8.355e-02 57.764 < 2e-16 ***
## monthaug -4.871e-01 1.018e-01 -4.786 1.70e-06 ***
## monthdec 8.613e-01 2.130e-01 4.044 5.25e-05 ***
## monthfeb -7.919e-02 1.247e-01 -0.635 0.525266
## monthjan -1.699e+00 1.814e-01 -9.363 < 2e-16 ***
## monthjul -8.772e-01 9.904e-02 -8.857 < 2e-16 ***
## monthjun 8.273e-01 1.214e-01 6.815 9.45e-12 ***
## monthmar 2.741e+00 1.847e-01 14.835 < 2e-16 ***
## monthmay -1.108e-01 9.160e-02 -1.209 0.226588
## monthnov -6.073e-01 1.039e-01 -5.846 5.04e-09 ***
## monthoct 1.671e+00 1.498e-01 11.156 < 2e-16 ***
## monthsep 1.190e+00 1.933e-01 6.157 7.41e-10 ***
## contactunknown -1.669e+00 8.637e-02 -19.323 < 2e-16 ***
## poutcomeother 6.015e-02 1.293e-01 0.465 0.641801
## poutcomesuccess 1.837e+00 1.284e-01 14.307 < 2e-16 ***
## poutcomeunknown -7.150e-01 1.354e-01 -5.281 1.28e-07 ***
## housingyes -5.076e-01 5.540e-02 -9.161 < 2e-16 ***
## age_categoricaMedianaEdad -3.433e-01 7.357e-02 -4.666 3.08e-06 ***
## age_categoricaMayores -3.929e-01 8.663e-02 -4.535 5.76e-06 ***
## age_categoricaAncianos 4.974e-01 1.455e-01 3.419 0.000629 ***
## educationsecondary 4.173e-02 7.754e-02 0.538 0.590411
## educationtertiary 3.712e-01 9.259e-02 4.009 6.10e-05 ***
## pdays -3.077e-03 5.098e-04 -6.036 1.58e-09 ***
## campaign_binaria+1Contacto -3.019e-01 4.784e-02 -6.310 2.79e-10 ***
## jobblue-collar -2.584e-01 8.881e-02 -2.910 0.003617 **
## jobentrepreneur -5.423e-01 1.572e-01 -3.450 0.000562 ***
## jobhousemaid -7.257e-01 1.763e-01 -4.115 3.87e-05 ***
## jobmanagement -2.481e-01 9.433e-02 -2.630 0.008529 **
## jobretired -2.328e-01 1.347e-01 -1.728 0.083902 .
## jobself-employed -4.134e-01 1.411e-01 -2.931 0.003380 **
## jobservices -2.470e-01 1.034e-01 -2.389 0.016897 *
## jobstudent 2.538e-01 1.527e-01 1.662 0.096551 .
## jobtechnician -1.785e-01 8.748e-02 -2.040 0.041355 *
## jobunemployed -3.030e-01 1.448e-01 -2.092 0.036432 *
## loanyes -2.630e-01 7.030e-02 -3.741 0.000183 ***
## maritalmarried -2.781e-01 7.185e-02 -3.871 0.000109 ***
## maritalsingle -1.306e-01 8.176e-02 -1.597 0.110198
## day 1.280e-02 3.273e-03 3.912 9.17e-05 ***
## balance 1.759e-05 6.565e-06 2.679 0.007377 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 21112 on 34107 degrees of freedom
## Residual deviance: 13348 on 34068 degrees of freedom
## AIC: 13428
##
## Number of Fisher Scoring iterations: 7</code></pre>
<div class="sourceCode" id="cb465"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb465-1"><a href="etl.html#cb465-1"></a>both_selection <-<span class="st"> </span><span class="kw">stepAIC</span>(empty_model, <span class="dt">trace=</span><span class="ot">FALSE</span>, <span class="dt">direction=</span><span class="st">"both"</span>, <span class="dt">scope=</span>horizonte)</span></code></pre></div>
<p>Con las tres aproximaciones se ha llegado al mismo resultado de las mismas 14 variables predictoras, por ello a la hora de realizar el modelado solo se va a utilizar un modelo.</p>
</div>
</div>
</section>
</div>
</div>
</div>
<a href="análisis-exploratorio-variables.html" class="navigation navigation-prev " aria-label="Previous page"><i class="fa fa-angle-left"></i></a>
<a href="modelado.html" class="navigation navigation-next " aria-label="Next page"><i class="fa fa-angle-right"></i></a>
</div>
</div>
<script src="book_assets/gitbook-2.6.7/js/app.min.js"></script>
<script src="book_assets/gitbook-2.6.7/js/lunr.js"></script>
<script src="book_assets/gitbook-2.6.7/js/clipboard.min.js"></script>
<script src="book_assets/gitbook-2.6.7/js/plugin-search.js"></script>
<script src="book_assets/gitbook-2.6.7/js/plugin-sharing.js"></script>
<script src="book_assets/gitbook-2.6.7/js/plugin-fontsettings.js"></script>
<script src="book_assets/gitbook-2.6.7/js/plugin-bookdown.js"></script>
<script src="book_assets/gitbook-2.6.7/js/jquery.highlight.js"></script>
<script src="book_assets/gitbook-2.6.7/js/plugin-clipboard.js"></script>
<script>
gitbook.require(["gitbook"], function(gitbook) {
gitbook.start({
"sharing": {
"github": false,
"facebook": true,
"twitter": true,
"linkedin": false,
"weibo": false,
"instapaper": false,
"vk": false,
"all": ["facebook", "twitter", "linkedin", "weibo", "instapaper"]
},
"fontsettings": {
"theme": "white",
"family": "sans",
"size": 2
},
"edit": {
"link": null,
"text": null
},
"history": {
"link": null,
"text": null
},
"view": {
"link": null,
"text": null
},
"download": null,
"toc": {
"collapse": "subsection"
},
"search": false
});
});
</script>
</body>
</html>