| Paper: | MLSP-P6.11 | ||
| Session: | Learning Theory and Models | ||
| Time: | Thursday, May 20, 15:30 - 17:30 | ||
| Presentation: | Poster | ||
| Topic: | Machine Learning for Signal Processing: Signal detection, Pattern Recognition and Classification | ||
| Title: | A GENERATIVE-DISCRIMINATIVE HYBRID FOR SEQUENTIAL DATA CLASSIFICATION | ||
| Authors: | Karim Abou-Moustafa; Concordia University | ||
| Ching Suen; Concordia University | |||
| Mohamed Cheriet; École de Technologie Supérieure | |||
| Abstract: | Classification of Sequential data using discriminative models such as SVMs, is very hard due to the variable length of this type of data. On the other hand, due to their efficiency generative models such as HMMs have become the standard tool for representing sequential data. This paper proposes a general generative-discriminative framework that uses HMMs to map the variable length sequential data into a fixed size P-dimensional vector (likelihood score) that can be easily classified using any discriminative model. The preliminary experiments of the framework on the MNIST database for handwritten digits haveachieved a better recognition rate of 98.02% than that of standard HMMs (94.19%). | ||
| Back | |||
Home -||-
Organizing Committee -||-
Technical Committee -||-
Technical Program -||-
Plenaries
Paper Submission -||-
Special Sessions -||-
ITT -||-
Paper Review -||-
Exhibits -||-
Tutorials
Information -||-
Registration -||-
Travel Insurance -||-
Housing -||-
Workshops