kernel   
Main
CS680
Instructor
 
Course Materials
Outline
Schedule
Homework
Projects
 
Resources
PyML
Python
  

detailed course schedule

Week Topic References
1 Preliminaries:
Course overview, biological preliminaries		 Introduction to molecular biology for computer scientists:
First chapter of ref. [1] [pdf]
Martin Tompa's notes [pdf]
2 Machine learning - a reminder
Demo of PyML [slides]
Examples of kernel methods
Kernels and kernel matrices		 Chapters 2 and 3 of ref. [2]
3 Kernels and kernel matrices (cont)
Classifier margin and large margin classification		 Section 2.4 of Chapter 2 in ref. [1] [pdf]
Chapter 7 of ref. [3]
4 Constrained convex optimization
Support Vector Machines 		Section 2.4 of Chapter 2 in ref. [1] [pdf]
Chapter 7 of ref. [3]
5 Support Vector Machines
Understanding the effect of SVM parameters
6 Closure properties of kernels
Constructing kernels
Simple operations in feature space.
Kernel K-means.
SVMs for unbalanced data.
Measuring classifier accuracy.
Section 3.4.1 in ref. [2]
Section 13.1 of ref. [3]
7 More on measuring classifier accuracy
SVMs for unbalanced data revisited
Guidelines for experiment design in machine learning.
The representer theorem.
Tips for giving good talks.
8 SVM training methods
Remote homology detection. 		Léon Bottou and Chih-Jen Lin. Support Vector Machine Solvers. [pdf].
9 Remote homology detection and the motif kernel
Convolution kernels and their applications
A. Ben-Hur and D. Brutlag. Remote homology: a motif-based approach [pdf].
Section 11.7 of ref. [2]
10 Kernel PCA and kernel CCA.
Fast kernels for strings: spectrum, mismatch and related kernels. 		Chapter 6 of ref. [2].
Chapter 4 of ref. [1] and related paper.
11 Recognition of alternative splicing with kernel methods.
The local alignment kernel. 		Chapter 13 of ref. [1]
RASE: Recognition of alternatively Spliced Exons in C. elegans.
chapter 6 of ref. [1] [pdf] and related paper
12 The Fisher kernel.
Weighted decomposition kernels. 		Chapter 12 of ref. [2]
T. Jaakkola, M. Diekhans, and D. Haussler. Using the Fisher kernel method to detect remote protein homologies [ps]
Weighted decomposition kernels. [pdf]
13 Feature Selection
Kernel methods for structured output spaces
Kernels for protein-protein interactions
Kernels on a hierarchy 		I. Guyone and A. Elisseeff. An Introduction to Variable and Feature Selection [pdf].
14 Kernel methods for structured output spaces (cont)
Kernel methods for protein-protein interactions
Kernels on a hierarchy 		A. Ben-Hur and W.S. Noble. Kernel methods for predicting protein-protein interactions [pdf].

Books:

  1. Bernhard Schoelkopf, Koji Tsuda and Jean-Philippe Vert. Kernel Methods in Computational Biology, MIT Press, 2004.
  2. John Shawe-Taylor and Nello Cristianini. Kernel methods for pattern analysis, Cambridge, 2004.
  3. Bernhard Schoelkopf and Alex Smola. Learning with Kernels, MIT press, 2002.