This course is based on class notes being written during the semester. Additional materials will be posted below as appropriate. Text in bold roman typeface is typically a link, so when a topic header turns bold you can click on it to access the notes. If a set of class notes is modified, the revision date next to it is updated accordingly.
This syllabus is a plan, not a commitment, and may change depending on class interest and the time needed to cover the various topics. Different notes may take very different amounts of time to cover. Any materials [in square brackets] in the Supplementary Materials column below, as well as the appendices in all the note sets, are optional reading.
Lecture Notes | Supplementary Materials |
---|---|
Overview and Logistics | |
Functions and Data Fitting (rev 9/1) | [s], [linear systems refresher] |
Introduction to Machine Learning (rev 9/5) | [s] |
Nearest Neighbor Predictors (rev 9/9) | [s] |
Function Optimization (rev 9/24) | [s] |
Linear Predictors (rev 9/25) | [s] |
Validation and Testing | [s] |
Decision Trees (rev 10/21) | [s] |
Random Decision Forests (rev 8/29) | [s] |
Convex Programs (rev 12/04) | [s] |
Support Vector Machines (rev 11/12) | [s] |
Kernels (rev 11/11) | [s] |
Convolutional Neural Nets (rev 11/26) | [s] |
Training Neural Nets by Back-Propagation (rev 11/26) | [s] |
Concluding Remarks: Topics Not Covered | [s] |
Status on Dec 6: Completed all notes above. Updated all scribble sets. |
COMPSCI 371D, Duke University, Site based on the fluid 960 grid system