Using a machine learning algorithm for a given application often requires tuning design parameters of the classifier to obtain optimal classification performance without overfitting. In this contribution, we present an evolutionary algorithm based approach for multi-objective optimization of the sensitivity and specificity of a -SVM. The -SVM is often preferred over the standard C-SVM due to smaller dynamic range of the parameter compared to the unlimited dynamic range of the C parameter. Instead of looking for a single optimization result, we look for a set of optimal solutions that lie along the Pareto optimality front. The traditional advantage of using the Pareto optimality is of course the flexibility to choose any of the solutions that lies on the Pareto optimality front. However, we show that simply maximizing sensitivity and specificity over the Pareto front leads to parameters that appear to be mathematically optimal yet still cause overfitting. We propose a multiple objective optimization approach with three objective functions to find additional parameter values that do not cause overfitting.
Abstract-Two important components of a speaker identification system are the feature extraction and the classification tasks. First, features must be robust to noise and they must also be able to provide discriminating information that the classifier can use to determine the speaker's identity. Second, the classifier must take the features that have been extracted from a sentence and label them as corresponding to one of the enrolled speakers. However, sets of features may be even more beneficial than any single feature by itself. There may be information present in one feature that other features do not have. Therefore, we present analysis of features and fusion by employing probabilistic averaging and weighted majority voting. Weighted voting will require that the weights are determined in a non-heuristic methodology and are robust to data with a large amount of channel distortion. Results using the King database show that both fusion methods lead to enhanced performance.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.