This repo provides python-based solutions to the Coursera Machine Learning course taught by professor Andrew NG.

Solutions mimic the MATLAB/Octave code provided through Coursera. The scikit-learn module provides canned implementations for each of these algorithms, but at the potential expense of understanding how they work. These basic versions are being used to solidify concepts and gain familiarity with the Python scientific computing stack.

Unlike in the coursera course, individual functions are not written in seperate scripts. Instead, a single script is included to store all of the helper functions for each exercise.

Exercise 1: Linear Regression (working)

Exercise 2: Logistic Regression (working)

Exercise 3: Multi-class Classification and Neural Networks (working)

Exercise 4: Neural Network Learning (working)

Exercise 5: Regularized Linear Regression and Bias/Variance (working)

Exercise 6: Support Vector Machines (not started)

Exercise 7: K-Means Clustering and PCA (working)

Exercise 8: Anomaly Detection and Recommender Systems (working)