We approached the problem of classifying papers for the TREC 2004 Genomics Track triage task as a four step process: feature generation, feature selection, classifier training, an...
Aaron M. Cohen, Ravi Teja Bhupatiraju, William R. ...
In typical classification tasks, we seek a function which assigns a label to a single object. Kernel-based approaches, such as support vector machines (SVMs), which maximize the ...
In this paper, we address two issues of long-standing interest in the reinforcement learning literature. First, what kinds of performance guarantees can be made for Q-learning aft...
A Natural Language Interface for Databases allows users of multimedia kiosks to formulate natural language questions. User questions are first translated into a logic language and ...
In this paper, westudy the application of an ttMM(hidden Markov model) to the problem of representing protein sequencesby a stochastic motif. Astochastic protein motif represents ...