Machine Learning Projects

Between teaching an introductory machine learning course and taking a course myself over the past semesters, I have completed several machine learning projects. The aim for each project was to understand the theory behind various machine learning techniques and the effects of various parameters in practice.

Skills:

Supervised learning methods
- Regression
- Classification
- Naïve Bayesian Model
- Decision trees
- SVM
- Weighted Knn
Unsupervised learning methods
- Clustering
- Neural Network

Featured Projects

Neural Network

By initializing a random weights then applying forward-propagation and backpropagation I trained a neural network. This chart compares training accuracies and testing accuracies as well as the effect of lambda (regularization parameter) and the number of epochs.

PCA Algorithm for Facial Recognition

With AT&T labs face dataset, the images are processed for PCA analysis. After extracting the mean face as well as the eigenvectors of the covariance matrix I used a weighted Knn and SVM classifier to create the model with ensemble learning.

K-means Clustering

Here I applied k-means clustering to various images to simplify the image with color. We compare how the number of clusters (k), the number of iterations, and repeat value effects the image result.

Regularization

Here I compare the average testing and training error with lambda to determine what best regularization parameter, lambda, to use in this situation for a regression model.