Information, Inference, and Learning (CS-0149)

Fall 2026 · CS 0149 · Tufts University · 3 credits · Undergraduate

Modern methods for inference and learning through the lens of information theory and statistics. Entropy, relative entropy (KL divergence), mutual information, information inequalities, and their operational interpretation in terms of data compression/rate-distortion. Rigorous bridge through large deviations and statistical inference to variational inference, latent variable models such as Variational Auto-Encoders (VAE), and modern diffusion-based generative models. The unifying theme is the interpretation of inference and learning as optimization of information-theoretic objectives, understructural or computational constraints.

Course codes: CS-0149, CS 0149, CS0149, CS-149, CS 149, CS149, EE-0127, EE 0127, EE0127, EE-127, EE 127, EE127

Course Discovery