My research interests include data mining, software metrics, software quality and reliability, software quality modeling, and bioinformatics. I am currently working on feature selection for software quality prediction.
We explore the data mining concept of feature selection
and investigate those technologies in the context of software
quality prediction and software metrics. Various techniques
developed from data mining and machine learning
have been successfully applied for deriving new information
in a variety of domains. Feature selection (or
attribute selection) has become a vital pre-processing step
in most data mining and machine learning problems. In
addition to improving the quality of the machine learning
data set, feature selection is particularly useful for highdimensional
data. The aim of feature selection
is to find a feature subset (i.e., data reduction) that can learn
and describe the data set such that it is equivalent to the
same task being done by the original data set (i.e., without
any data reduction).
Chair, thesis commettee:
Member, thesis commettee:
- Pengpeng Lin
Title: "A Framework for Consistency Rate Based Feature Selection"
Status: completed in May 2009
Pengpeng is a Ph.D. student in the Department of Computer Science at the University of Kentucky now.
- Sri Harsha Vege
Title: "A Study on Feature Selection Techniques in Bioinformatics"
Status: expect to graduate in May 2012
- Junhua Wang
Title: "Large-Sample Logistic Regression with Latent Covariates in a Bayesian Networking Context"
Status: completed in July 2009
- Prapanna Tamarapu Parthasarathy
Title: "Integrated Component Metrics and their Evaluation"
Status: completed in May 2007