Selected Features for Classifying Environmental Audio Data with Random Forest

Environmental audio classification has been the focus in the field of speech recognition. For Environmental audio data, it is difficult to find an optimal classifier and select the optimal features from various features can be extracted. Random forest is a powerful machine learning classifier compared to other conventional pattern recognition techniques. In this paper, the performance of the Random Forest classifier and the selected features model for environmental audio classification are explored. The comparison and analysis of classification results are given which obtain by employing the Bagging, AdaBoost, and Random Forest for environmental audio data. The selection of importance variables in building the classification model and assessment variable importance are involved in the experiments. The experimental results show that the Random Forest method can effectively improve the performance of environmental audio data classification even under the fewer number of the training examples. According to the variable importance, the model built to improve on both the efficiency and the accuracy of classification based on the selected features in environmental audio data.