Assignment Descriptions
All written assignments are to be done with a word processor and be neat and professional. Good writing, grammar, punctuation, etc. are important and can affect your grade.
Wiki Entry
Sept 7, 2007
Do some surfing on the web and find at least two examples of recent, interesting machine learning news, applications, advances, etc. Make an entry on the course wiki that briefly describes each, gives your opinions, etc. and points interested readers to more information.
Learning System
Sept 14, 2007
In not more than two typed pages, propose and describe a system that you can argue learns, or improves its behavior over time. Show examples of its behavior and convince me that it displays some ability to learn. Be creative. Bonus points if your system is fun.
Paper Presentations
Sept 19–Oct 15, 2007
In 5-10 minutes give an overview of your chosen research paper. Make sure to clearly state the paper's significance and support it with a high-level description of the approach taken and results achieved. In other words, tell us briefly what the goal of the paper was, how they attempted to achieve the goal and how successful they were. Be prepared to answer questions.
Some good sources for potential papers include (but are not limited to):
- Proceedings of Neural Information Processing Systems
- Proceedings of the International Conference on Machine Learning
- Neural Networks
- IEEE Transactions on Neural Networks
- Neural Computation
- Machine Learning
- Journal of Artificial Intelligence Research
- Journal of Machine Learning Research
Presentation Schedule
- Sep 24 — Mike Smith
- Outlier Detection by Active Learning
- Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. ???, 2006.
- Sep 26 — Yan Shi
- Pareto-based Multiobjective Machine Learning: An Overview and Case Studies
- IEEE Transactions on Systems, Man and Cybernetics (Part C) 38(3), pp. ???, 2008.
- Oct 01 — Jie Long
- Gesture-Based Affective Computing on Motion Capture Data
- Proceedings of the International Conference on Affective Computing and Intelligent Interaction, pp. ???, 2005.
- Oct 03 — Aaron Dennis
- Machine Learning for Sequential Data: A Review
- Structural, Syntactic, and Statistical Pattern Recognition, pp. 227-246, 2002.
- Oct 08 — Heather Chan
- Cascaded Bidirectional Recurrent Neural Networks for Protein Secondary Structure Prediction
- IEEE/ACM Transactions on Computational Biology and Bioinformatics 4(4), pp. 572-582 2007.
Topic Proposal
Oct 5, 2007
Do some outside reading, thinking, research, etc. and come up with a solid topic for your research project. Write a paragraph (up to half a page) describing your proposal, why it is important and how you will approach it. I will review these, make suggestions and consult with you in the following week to make sure you have a solid topic, appropriate in scope for the course and the time available.
High-level Potential Project Areas:
- Variations/extensions of any of the course topics
- Development of a new algorithm/approach to learning
- Interesting application/hybrid approach to your field of research
Project Abstract
Nov 28, 2007
Turn in a well-written abstract describing your class project. Include a descriptive title, description of the work and intended results (around 200 words), and an annotated bibliography that includes at least five references from the literature. Please follow this format.
Written Project
Dec 12, 2007
A major part of this course will involve your development of a class project. This will involve significant outside study and preparation on your part and will consist of both a written paper and an oral presentation to the class. Ideally, the paper can evolve into a publishable paper and the oral presentation will give you some experience in presenting your research before peers. An example of an acceptable project is to thoroughly research and present an overview of a topic we do not cover in class. An example of a better project is to attempt some modification, extension or improvement to existing work. The best type of project involves your best efforts at producing some novel results (new algorithm, theoretical results, etc.) The paper should be well-written and professionally presented as if you were going to submit it for publication. A schedule of the oral presentations is here.