VOICE RECOGNITION USING CLASSIFICATION ALGORITHMS

In Speech recognition systems are based on machine learning methods, among which classification algorithms are widely used. Classification performs the task of dividing voice signals into various categories, such as words or sentences. Commonly used algorithms include logistic regression, decision trees, and neural networks. During voice signal processing, features, i.e., important parameters, are first extracted and then passed to the classifier. Based on the classification results, the system converts speech into text or determines the specific content of the sound. This technology is essential for improving human-computer interaction.This article discusses a classification algorithm for the problem of speech identification using the machine learning method. The MFCC algorithm is used for preprocessing speech. To solve this problem, a comparative analysis of five classification algorithms was carried out. In the first experiment, the methods of the reference vector – 0.90 and the multilayer perceptron – 0.83 were determined and the best results were shown. In the second experiment, a multilayer perceptron with an accuracy of 0.93 was proposed using the Robust Scaler method for personality identification. Therefore, a multilayer perceptron can be used to solve this