List view

• Jann-0.5

  Support Vector Machines --------------------------------- Implement Support vector Machines Unsupervised Learning -------------------------------- K-Means Algorithm Dimensionality Reduction ------------------------------------ (PCA) Principal Component Analysis Notes and Logic behind applying PCA Anomaly Detection ------------------------- Recommender Systems ---------------------------------

  Overdue by 13 year(s)

  •

  Due by August 31, 2012
  0% complete0 open0 closed

• Jann-0.4

  Part 1 --------- Linear regression One/Multiple Variables Linear regression Training: Batch GD (single neuron) Stochastic GD (single or nNeurons) Feature Mapping with degrees Feature Scaling with (STD, average) Normal Equation (single neuron) Dynamic Training through com.marmoush.jann.train.Train Class Better Tests *** Backpropagation will be deleted and replaced with better more clear algorithms Part 2 -------- Models: 1- Logistic Regression One Class 2- Logistic Regression Multiple Classes Training: Logistic Regression GD Regularized Linear Regression Regularized Logistic Regression Regularized Normal Equation Utils: Random Initialization of theta Part 3 ------- Models: 1- Logistic Regression Neural Network 2- Linear Regression Neural Network Gradient Checking Numerical Estimation of gradients Neural Network Back propagation Part 4 -------- Notes or Logic for Decision Making For handling errors : 1-Identify Problems with the help of (Graphs, tools) 2-Act upon on it (Manual ways, Automatic ways) Part 5 -------- More Tests More Graphs

  Overdue by 13 year(s)

  •

  Due by July 31, 2012

  •13/18 issues closed
  72% complete5 open13 closed