Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

Minimum Bayes Error Feature Selection

George Saon, Mukund Padmanabhan

IBM T. J. Watson Research Center, Yorktown Heights, NY, USA

We consider the problem of designing a linear transformation Θ ∈ Rpxn, of rank p ≤ n, which projects the features of a classifier x ∈ Rn onto y = Θx ∈ p such as to achieve minimum Bayes error (or probability of misclassification). Two avenues will be explored: the first is to maximize the Θ-average divergence between the class densities and the second is to minimize the union Bhattacharyya bound in the range of Θ. While both approaches yield similar performance in practice, they outperform standard LDA features and show a 10% relative improvement in the word error rate over state-of-the-art cepstral features on a large vocabulary telephony speech recognition task.


Full Paper

Bibliographic reference.  Saon, George / Padmanabhan, Mukund (2000): "Minimum Bayes error feature selection", In ICSLP-2000, vol.3, 75-78.