11-755 · Fall 2026
Schedule
Fall 2026 calendar to be announced
TBA
The Fall 2026 schedule has not been released yet
Lecture dates, slides, recordings and homework releases will be published here once the university timetable is confirmed. The outline below shows the material we expect to cover; the order and the exact set of topics are subject to change based on student interests and course discussions.
Lecture slides from previous editions are not linked while the calendar is being rebuilt. Registered students will find everything on Canvas and Piazza once the term begins.
What the course covers
Grouped by theme — not the lecture order.
-
Module 1
Foundations
- Introduction to machine learning for signal processing
- Linear algebra for signals, part 1
- Linear algebra for signals, part 2
- Optimization
- Probability and information theory
-
Module 2
Representations
- Deterministic representations
- Data-driven representations
- Non-negative matrix factorization
- Sparse and overcomplete representations
- Dictionary learning
-
Module 3
Learning from signals
- Boosting and face detection
- Clustering
- Regression and prediction
- Linear classifiers
-
Module 4
Latent variable models
- Independent component analysis
- Canonical correlation analysis
- Expectation maximization, part 1
- Expectation maximization, part 2
- Factor analysis
- Maximum entropy modeling
-
Module 5
Sequences and state-space models
- Hidden Markov models, part 1
- Hidden Markov models, part 2
- Kalman filtering and tracking
- Extended and non-linear filtering
-
Module 6
Applications and special topics
- Survival analysis
- Predicting rare events
- Course project presentations
Coursework
Weights are described in the syllabus.
- Homeworks Four assignments Dates TBD Released across the semester as mini-projects. Handouts are posted on Canvas.
- Quizzes Weekly Dates TBD Short multiple-choice quizzes on the previous week's material.
- Group project Team project Dates TBD Proposal, midway report and final presentation, scheduled once the term calendar is set.