Principles of Machine Learning in Voice Recognition

Sounds carry a huge amount of information of our daily life and physical events. We hear sounds in complex environment and from increased amounts of electric devices, so audio recognition is definitely an important technology in human-computer interaction. In recent years, speech recognition technology developed rapidly. Summarizing recognition methods is necessary when there are too many methods to choose. This paper introduces some of the advanced technologies, Recurrent Neural Network, Convolutional Neural Network, Linear Regression, Multilayer Perceptron, Random Forest and their principle.