Computer Science Colloquium Series
Pham Quang Huy
Date: Friday, March 3rd, 2017
Time: 11:00 am
Location: Chrysler Hall North, G100
Abstract: Feature selection is a very important task in many machine-learning problems to yield more compact data, thus, ensuring a chosen model applicable for the domains involving high dimensional data. In some cases, it may even increase the classification/regression quality. Many feature selection algorithms have been proposed, but they have their own drawbacks, e.g., some are too greedy while some other are too exhaustive. On the other hand, parameter search is another factor to improve the performance of the model. However, if it is applied after the feature selection step, the classification performance might not be as optimal as it should. We proposed a method, called AFS, which integrates the parameter search in the feature selection process to avoid the mentioned drawbacks. Furthermore, its greedy/exhaustive behavior can be controlled by the user so that it is flexible to the problem of interest. In our experiments on several real-life datasets, including biological data, AFS often produces better results than many complicated combinations of effective feature selection approaches and classifiers available in Weka, a very common powerful software tool for machine learning tasks. It can be a good choice for biological classification/prediction problems that contain a lot more features than samples.
Bio: Pham Quang Huy received his Bachelor’s degree in computer science from the University of Dalat, Vietnam, in 2000, and his Master’s degrees in computer science from the Natural Science University of Ho Chi Minh City, in 2005. He is currently a PhD student in Computer Science at the University of Windsor. His research interests are mainly focused on data-mining, machine learning and pattern recognition, mostly in the fields of closed frequent item-set mining, association rules mining, chemotherapy response prediction and drug-target interaction prediction.