Assignments

Problem Sets

There will be one math problem set (PS) for each topical unit of the course.

These are intended to build your skills in formal mathematics (analysis, derivation).

• PS0: Probability Fundamentals
• PS1: Beta, Dirichlet, and Estimators of Unigram Probability
• PS2: Gaussians and Estimators
• PS3: KL Divergence, ELBO, and Variational Inference
• PS4: K-means and Gaussian Mixtures

Coding Homeworks

There will be one coding assignment (CA) for each topical unit of the course.

These are intended to build your implementation skills in Python, and will often be tightly intertwined with the same unit's math homework.

• CA1: Probabilistic Prediction of Words
• CA2: Bayesian Linear Regression
• CA3: Variational Autoencoders
• CA4: Coordinate Ascent for Gaussian Mixtures

Project

In the last half of the course, you'll complete a open-ended team project.

See our Project Overview page.

• Deliverable 1: Team Formation
• Deliverable 2: Project Pitch
• Deliverable 3: Chat with Course Staff
• Deliverable 4: Initial Report
• Deliverable 5: Final Report

Exams

In-class exams will be given after each unit of the course.

These are primarily designed to verify your mathematical and conceptual understanding. You can view each homework as good practice for the corresponding exam.

See the Schedule for the scheduled dates of each exam.