CSE 254: Seminar on Learning Algorithms

Time
TuTh 9.30-11 in CSE 2154

Instructor:
Sanjoy Dasgupta
Office hours Th 11-1 in CSE 4138

This quarter the theme of CSE 254 is neurally-inspired unsupervised learning.

The first few lectures will cover basic technical material. Thereafter, each class meeting will consist of student presentations. Each student will present a technical paper (or several papers) in detail. All seminar participants will discuss the paper and the issues raised by it.


Date	Presenter	Paper/Topic	Slides
Sep 27	Sanjoy	Random projection
Oct 2	Sanjoy	Hopfield nets
Oct 4	Sanjoy	Markov random fields
Oct 9	Sanjoy	Gibbs sampling and restricted Boltzmann machines
Oct 11	Sanjoy	Sparse recovery
Oct 16	Dalin	Learning a sparse code for natural images	slides
	Huanqiu	Efficient auditory coding and Efficient coding of natural sounds	slides
Oct 18	Jiayuan	Distributed neurally-plausible sparse coding	slides
	Shilin	Natural image statistics and neural representation	slides
Oct 23	Gokce	The independent components of natural scenes are edge filters	slides
	Vijay	The neocognitron	slides
Oct 25	Alankrita	Compressed sensing	slides
	Gaurav	More compressed sensing	slides
Oct 30	Shuang	Upper bounds on the capacity of Hopfield nets	slides
	Sophia	Expander codes	slides
Nov 1	Yaoguang	Properties of restricted Boltzmann machines	slides
	Zhiwei	Deep belief nets	slides
Nov 6	Jongha	Mean shift for gradient estimation and mode finding	slides
	Siva	Consistency of mean shift	slides
Nov 8	Jenny	Self-organizing maps	slides
	Chaitanya, Henry	Sparseness and expansion in sensory representations	slides
Nov 13	Max	Memorization and association on a realistic neural model	slides
	Robi	Unsupervised learning through preidction in a model of cortex	slides
Nov 15	Andrew, Anthony	Probabilistic logics and the synthesis of reliable organisms from unreliable components	slides
	Ivan	Highly fault-tolerant parallel computation	slides
Nov 20	Julaiti	Information bottleneck theory of deep learning: part 1 and part 2	slides
	Fangyi	Manifold learning via Isomap: overview and theory	slides
Dec 14	Duong	Laplacian eigenmaps	slides

This is a four unit course in which the work consists of oral presentations.

The procedure for each student presentation is as follows:

One week in advance: Finish a draft of Latex/Powerpoint that present clearly the work in the paper. Make an appointment with me to discuss the draft slides. And email me the slides.
Several days in advance: Meet for about one hour to discuss improving the slides, and how to give a good presentation.
Day of presentation: Give a good presentation with confidence, enthusiasm, and clarity.
Less than three days afterwards: Make changes to the slides suggested by the class discussion, and email me the slides in PDF, two slides per page, for publishing. Try to make your PDF file less than one megabyte.

Please read, reflect upon, and follow these presentation guidelines, courtesy of Prof Charles Elkan. Presentations will be evaluated, in a friendly way but with high standards, using this feedback form.

The schedule of presentations will be determined as much as possible on Sep 27. Here is a preliminary list of papers.

If you want to change your presentation date, please arrange a swap with another student and notify me at least two weeks in advance.