The idea of representing images using a bag of visual words is currently popular in object category recognition. Since this representation is typically constructed using unsupervi...
Abstract-- As intrusion detection essentially can be formulated as a binary classification problem, it thus can be solved by an effective classification technique-Support Vector Ma...
- Metagenomics is the study of environmental samples. Because few tools exist for metagenomic analysis, a natural step has been to utilize the popular homology tool, BLAST, to sear...
Background: Profile Hidden Markov Models (HMM) are statistical representations of protein families derived from patterns of sequence conservation in multiple alignments and have b...
Prashant K. Srivastava, Dhwani K. Desai, Soumyadee...
We describe an acoustic chord transcription system that uses symbolic data to train hidden Markov models and gives best-of-class frame-level recognition results. We avoid the extre...