INTERSPEECH 2007

Kernel logistic regression (KLR) is a popular nonlinear classification technique. Unlike an empirical risk minimization approach such as employed by Support Vector Machines (SVMs), KLR yields probabilistic outcomes based on a maximum likelihood argument which are particularly important in speech recognition. Different from other KLR implementations we use a Nyström approximation to solve large scale problems with estimation in the primal space such as done in fixedsize Least Squares Support Vector Machines (LSSVMs). In the speech experiments it is investigated how a natural KLR extension to multiclass classification compares to binary KLR models coupled via a oneversusone coding scheme. Moreover, a comparison to SVMs is made.
Bibliographic reference. Karsmakers, Peter / Pelckmans, Kristiaan / Suykens, Johan / hamme, Hugo Van (2007): "Fixedsize kernel logistic regression for phoneme classification", In INTERSPEECH2007, 7881.