credit-cards-ppt.Rmd

---
title: "Maching Learning in R using mlR"
author: "Surag Gupta"
date: "August 20, 2017"
output: beamer_presentation
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE,
                      cache = TRUE)
```

## About me

- Senior Analyst, Assortment Optimization
- MS Business Analytics, University of Minnesota (Class of 2017)
  - which means I'm a relatively new team member
- Worked as a Decision Scientist for 3 years prior to my graduate program
- R Enthusiast (This would be a good time for you to notice my t-shirt, if you haven't already)

## Disclaimer

Those who use Python for data science tasks shall not throw stuff at me during the session

## What is mlR?

mlR is an R package developed to simplify basic and advanced data science tasks, such as:

- Data preparation
- Feature Engineering
- Training Models
- Validating models
- Predictions
- ...

## We use the Kaggle credit card dataset to demonstrate the use of `mlR`



## Load packages

```{r cars, echo = FALSE, message = FALSE}

required_packages <- c( 
  # Add to this list the packages that you will use - if unavailable, it will be 
  # automatically installed
    "Hmisc",
    "caret",
    "kknn",
    "gbm",
    "nnet",
    "h2o",
    "xgboost",
    "lightgbm",
    "mlr",
    "parallel",
    "parallelMap",
    "ggplot2",
    "data.table",
    "tidyverse"
    )

  packages_to_install = required_packages[!(required_packages %in% 
                                                installed.packages()[, 1])]
    
  if (length(packages_to_install) > 0) {
    install.packages(packages_to_install)
  }
    
  suppressPackageStartupMessages({
    sapply(required_packages, require, character.only = TRUE)
  })
```

## Import the dataset

```{r message = FALSE, results = "hide"}
#Use fread() from the data.table package to read
df <- fread("creditcard.csv")  
                              
df <- df %>% 
  data.frame() %>% 
  mutate(Class = as.factor(Class)) #Convert target variable to categorical type
```

```{r}
head(df)
```