Title :
Multi-Class Classification Using a New Sigmoid Loss Function for Minimum Classification Error (MCE)
Author :
Ratnagiri, M.V. ; Rabiner, L. ; Biing-Hwang Juang
Author_Institution :
Dept. of Electr. & Comput. Eng., Rutgers Univ., Piscataway, NJ, USA
Abstract :
A new loss function has been introduced for Minimum Classification Error, that approaches optimal Bayes´ risk and also gives an improvement in performance over standard MCE systems when evaluated on the Aurora connected digits database.
Keywords :
Bayes methods; pattern classification; Aurora connected digits database; minimum classification error; multi-class classification; optimal Bayes risk and; sigmoid loss function; Accuracy; Databases; Estimation; Hidden Markov models; Kernel; Loss measurement; Noise; Bayes Risk; Minimum Classification Error; Savage Loss; sigmoid loss;
Conference_Titel :
Machine Learning and Applications (ICMLA), 2010 Ninth International Conference on
Conference_Location :
Washington, DC
Print_ISBN :
978-1-4244-9211-4
DOI :
10.1109/ICMLA.2010.20
