This repo contains Julia Rodd's assignments from the MSDS 420 Machine Learning class. This class was taught exclusively in Python.
The data folder contains data for most assignments. The data for assignments 5-8 are not directly shared due to their large size.
The output folder contains Jupyter notebooks. Both the code and write-up are provided within these files.
Located within the output folder, the assignments can be summarized as follows:
- Assignment 2: uses data from past bank marketing campaigns to predict responses. Logistic regression, Naive Bayes, and Random Forests are three methods considered.
- Assignments 3-4: uses Boston housing market data to predict home values. Regression-based methods (linear, lasso, ridge) and tree-based methods (random forests and gbms) are used.
- Assignment 5: uses MNIST data to demonstrate PCA and its effectiveness using a Random Forest classifier. Results are submitted to Kaggle.
- Assignments 6-8: these assignments comprise the deep learning portion of the class. Assignment 6 is an introduction to deep learning using sklearn MLPClassifier, while assignments 7 (CNN) and 8 (RNN) utilize tensorflow.
- Julia Rodd