CIS 700: Machine Learning for Bioinformatics
This course will explore the application of machine learning to open
problems in genomics, proteomics and related areas. The course will
be organized around the selection, specification, execution, and
presentation of research projects. Students will work in teams
to identify, define, and carry out a research project leading to a potentially
publishable paper. Background knowledge, techniques, and research
strategies will be discussed in class.
- Instructor: Lyle Ungar, ungar@cis.upenn.edu
- Prerequisites: CIS/GCB/Bio 536 Computational Biology or cis/gcb535 or permission of instructor
- This is not a lecture course, and is not appropriate for students with no background in biology.
Schedule
Example projects
- Effect of proximity of genes on the chromosome on gene expression
or replication. Develop probabilistic models which incorporate
both proximity and other data to model experiments.
- Protein function prediction using features selected from tens of
thousands of potential features automatically extracted from
databases. What features might be useful? How can we tell which ones
really are?
- Automatic generation of gene regulation networks by combining prior
literature knowledge with gene expression data. What can (or can't)
realistically be learned from real data?
previous year's web sites
return home
ungar@cis.upenn.edu