News Archive

2006

• Subramani Mani joined DSL lab. Dr. Mani holds a Ph.D. from the Intelligent Systems Studies Program of the University of Pittsburgh and specializes in Computational Causal Discovery and other Machine Learning methods. He will be working on new methods development as well as working with biomedical researchers on campus for solving real life modeling and data analysis problems using cutting edge DSL and broader Machine Learning methods and software. Welcome Dr Mani!
• "The Max-Min Hill-Climbing Bayesian Network Structure Learning Algorithm" (by I. Tsamardinos, L.E. Brown, C.F. Aliferis) has been published in Machine Learning, 2006.

2005

• GEMS poster at ISMB 2005 (by Statnikov, Tsamardinos, and Aliferis) is a winner of the best poster competition. See this link for details.
• Yindalon Aphinyanaphongs was awarded the Donald Lindberg Research Fellowship. The fellowship supports research to build and evaluate a novel medical information retrieval system for answering clinical questions.
• Jane McLaughlin wins The Third Annual People's Choice Award for Integrity, Respect, Pioneering, Service, and Collaboration. Congradulations, Jane!

2004

• Alexander Statnikov receives Gold Medal in the MedInfo 2004 Student Paper Competition for his microarray based cancer diagnosis work [Article].
• Machine Learning for Biomedical Decision Support and Discovery (Tutorial, presented at MedInfo 2004).

2003

• Yin Aphinyanaphongs receives First Prize in the 2003 AMIA Student Paper Competition for his text categorization work: [Article].
• HITON, A Novel Markov Blanket Algorithm for Optimal Variable Selection: [Article].
• A small-sample algorithm to find local causal relationships and Markov Blankets in a very high dimensional data: [Article]
.
• Scaling-Up Bayesian Network Learning to Thousands of Variables Using Local Learning Technique: [Article].
• A practical way to infer Markov Blankets using standard decision tree software: [Article].
• Causal Explorer: A Probabilistic Network Learning Toolkit for Biomedical Discovery is available for download, also see the following [Article].
• Robustness and inductive bias of feature selection methods in gene expression data: [Article], [Article].
• Gene expression data enables near-perfect classification of Lung Cancer types: [Article], [Article].
• Cleaning the haze of filters and wrappers: [Article]
.

2002

• DSL featured in peer review.
• New algorithms for efficient induction of Markov Blankets and local causal neighborhoods: [Article], [Article].