MLS 2012 - Machine Learning for Sequences 2012

Sequence classification and sequence labeling is at the heart of many pattern recognition and data mining tasks, in fields such as speech and handwriting recognition, bioinformatics, etc. Beyond well known Hidden Markov Models (HMMs), which have been widely used for modeling sequences of patterns, a number of alternative methods and models have been proposed in the recent years. These approaches include for instance discriminative training (e.g. large margin) of Hidden Markov Models, on-line learning of such models, discriminative models based on Conditional Random Fields, etc.
This special session aims at sharing new ideas and works on models and approaches for improving over state of the art methods for signal and sequence classification and labeling tasks.