General Section
Overview
Program Administration
People
Faculty Trainers
Predoctoral Trainees
Postdoctoral Fellows
Past Trainees
Events
CIBM Seminar Series
Annual Training Meeting
Annual Retreat
Presentations
News
Ph.D. Program Links
Program Details
Graduate
Postdoctorate
Short-Term Research
|
 |
Predoctoral Trainee
 |
Irene M. Ong
Computer Sciences Doctoral Training Program
Department of Computer Sciences
Faculty Supervisor: C. David Page, Jr.
Email: ong@cs.wisc.edu
263-7626
|
My research interests lies in data mining, particularly
Bayesian statistical methods, machine learning and inductive logic programming
with a focus on bioinformatics problems. I am particularly interested
in gene expression analysis and inferring biological networks, specifically,
inferring how genes regulate other sets of genes in the bacterium E.
coli from time series gene expression array data.
Our previous work showed that we could learn that expression of a gene
at one time step is a consistently good predictor of another gene at the
next time step given gene expression data and an operon map of E. coli.
The dynamic Bayesian network (DBN) framework is a probabilistic graphical
framework that can visually and intuitively represent time and the network
of regulation. However, the task of learning the graphical structure of
the DBN, which amounts to searching over the space of possible graph structures
for the most likely graph given the data, is computationally expensive.
We used prior knowledge in the form of the operon map to simplify our
learning task by grouping genes that are already known to be expressed
together into operons, thus reducing the search space. I am currently
working on ways to further improve the accuracy of the model.
CV
I Ong CV (http://www.cs.wisc.edu/~ong/resume.pdf)
Publications
Search for publications by name (Pub
Med, Cite
Seer)
|
