ECE521_Assignments

This repository documents the machine learning and data visualization assignments done as part of ECE521 coursework.

Assignments

ECE521 assignments on machine learning. Assignments are done in teams of two. Below are brief descriptions for each assignment.

Assignment 1

• Predicting values using k-nearest neighbours
• "3/5" digit classifier using linear regression and logistic regression
• Stochastic gradient descent
• Curse of dimensionality
• Bias-variance tradeoff
• Regularization

Assignment 2

• Multiclass classification of characters A to J using logistic regression
• Multiclass classification of characters A to J using neural networks
• Maximum likelihood estimation
• Early-stopping
• Dropout
• Softmax
• Hinton diagrams

Assignment 3

• Clustering using K-means
• Clustering using Mixture of Gaussians
• Discovering latent dimensions using factor analysis

Assignment 4

• Bayesian networks
• Factor graphs
• Hidden markov models
• Message-passing algorithm