The following repository contains Kenny's personal machine learning projects and experiments.
Project 1 contains a two part project that explores models to predict customer churn. Part A is an interpretable model, namely, a decision tree model, and Part B explores a black box model using emsembling (Random forests and AdaBoost). Findings are presented in a report which can be found in the folder.
Project 2 explores a multi-layered-perceptron model for prediction Parkinson's diseases from speech data. It contains an in-depth consideration of hyperameters including optimzers, number of neurons, number of hidden layers etc. Findings are presented in a report which can be found in the folder.
Misc contains misecellaenous code for dabblings with ML concepts and ideas.