Your class project is an opportunity for you to explore an interesting statistics problem of your choice with the tools and methods related to nonparametric regression and classification. You can choose from one of the three options: (1) a list of projects provided, (2) recreate results from a paper listed, or (3) propose your own idea.
It will be an individual project. Your project will be worth 40% of your final class grade, and will have two final deliverables:
-
a writeup in the format of a NIPS paper (8 pages maximum in NIPS format, including references; this page limit is strict), due on Tuesday June 10 (emailed to the instructors list, NO LATE SUBMISSION ACCEPTED since we need to get your grades in), worth 60% of the project grade.
-
a poster presenting your work for the class poster session on Friday June 6 (Time TBA) in the Atrium of the Allen Center (CSE), worth 20% of the project grade (Please see the details in the Poster Session below).
In addition, you need to turn in a midway progress report (3 pages maximum in NIPS format, including references) describing the results of your first experiments by Thursday May 22, worth 20% of the project grade. Note that, as with any conference, the page limits are strict! Papers over the limit will not be considered.
Please turn in a brief project proposal (1-page
maximum) by Thursday April
24. Read the list of available data sets and potential
project ideas below.
If you prefer to use your own data set, we will consider your proposal,
but you must have
access to this data already, and present a clear proposal for what you
would do with it.
Project proposal format: Proposals should be one page maximum. Include the following information:
-
Project title
-
Data set
-
Project idea. This should be approximately two paragraphs.
-
Papers to read. Include 1-3 relevant papers. You will probably want to read at least one of them before submitting your proposal.
-
Milestone: What will you complete by the milestone? Experimental results of some kind are expected here.
Here are some details on the poster session.
- We will provide poster boards that are 32x40. The two most common
ways to make your poster are:
- You can create a huge slide (in powerpoint, using beamer, etc) and print it out at one of the poster printers available.
- You can create a bunch of "normal" presentation slides, print out each one on a piece of (letter-sized) paper, and put them all together on a poster board. This is somewhat messier (and not as pretty, but we don't mind) and you don't need to wait for the special poster printer to print.
- Please arrive 15mins early to set up.
- We will provide easels, pins, and posterboards sized 32x40.
- Make sure your poster prominently displays the title and your name.
- During the poster session both Amrit and I will see your project. Please let us know if that doesn't happen.
- Given the large number of projects, we cannot spend a lot of time in each poster. Please prepare a *3 minute* spiel on your project. Think about it as a poster spotlight at a conference (3 minutes + questions)..
- We expect everyone to be there for the whole time, so we can see your work, you can see each others work and other people at UW can also see the cool things you have done.
Advanced Readings / Project Ideas
Piecewise polynomial
- Ryan J. Tibshirani (2013) Adaptive Piecewise Polynomial Estimation via Trend Filtering.
Gaussian processes
- Duvenaud et al (2011) Additive and projective models; Additive Gaussian Processes.
- Gilboa et al (2012) Additive and projective models; Structured Gaussian Processes.
- E.B. Fox and D.B. Dunson (2012) Partition models; Multiresolution Gaussian Processes.
- Robert B. Gramacy and Herbert K. H. Lee (2012) Partition models; Bayesian Treed Gaussian Process Models With an Application to Computer Modeling.
- Adams et al (2009) Density estimation; The Gaussian process density sampler.
- Edward Meeds and Simon Osindero (2006) Mixture of experts; An alternative infinite mixture of Gaussian process experts.
- Any paper listed here Others, including GP classification.
- Iwata et al (2013) Mixtures of GPs for flexible clustering.
Inference techniques for Dirichlet processes mixture models
- Kalli et al (2011) Slice sampling mixture models.
- David M. Blei and Michael I. Jordan (2006) Variational inference for Dirichlet process mixtures.
- Matt Hoffman, David M. Blei, Chong Wang, John Paisley (2012) Stochastic Variational Inference.
Others
- Chipman et al (2008) Bayesian Additive Regression Trees (BART).