Feature Selection Using Maximum Feature Tree Embedded with Mutual Information and Coefficient of Variation for Bird Sound Classification
Algorithm 1
MICV-ERMFT feature selection.
Name: MICV-ERMFT feature selection.
Input: Dataset (: number of samples; : number of features).
Step:(1)Calculate using equation (5) for each feature in (2)Sort the feature sequence in ascending order to obtain . According to , select data in and gradually add one, and use the base classifier to score. Delete the feature that led to the decline of the index, and obtain the feature sequence , and map to to get dataset .(3)Calculate Pearson correlation coefficient matrix for the feature vector by .(4)Apply algorithm BMFT (Algorithm 2) to construct a maximum feature tree for .(5)Apply two-neighborhood based redundancy eliminating algorithm ERFTN (Algorithm 3) on ; denote the result array as .(6)Map to to get dataset .
