Machine Learning Jupyter Notebook Exercises

Overview

This is a collection of several machine learning exercies I created as a teaching assistant for Professor David Anastasiu at Santa Clara University.
Designed for graduate course COEN 240 - Machine Learning
The notebooks run out of the box giving a tutorial of each algorithm while addressing questions like

• How does the math behind this work?
• When should I employ this?
• How do you adjust hyperparameters for better performance?

SVM Notebook

This activity demonstrates using support vector machines (SVMs) with various example 2D datasets. The data is purposely two dimensional so that plots with the decision boundary can be visualized.

Libraries: sklearn, pandas, numpy, matplotlib
Data: Adapted from a Stanford Machine Learning course [https://github.com/syaning/stanford-machine-learning/tree/master/machine-learning-ex6]

LDA Notebook

This activity demonstrates using linear discriminant analysis (LDA) for multi-class classification

Libraries: sklearn, pandas, numpy, matplotlib, scipy
Data: Iris dataset from UCI Machine Learning Repository [https://archive.ics.uci.edu/ml/datasets/Iris]

Ensemble Methods Notebook

This activity demonstrates two different types of ensemble methods, bagging and boosting

Libraries: sklearn, pandas, numpy, matplotlib, scipy
Data: Sonar dataset from UCI Machine Learning Repository [https://archive.ics.uci.edu/ml/machine-learning-databases/undocumented/connectionist-bench/sonar/]