Research on feature selection algorithms based on random forest

In recent years, the rapid development and application of social network, image recognition, biological information technology has produced many types of high dimensional datasets. Before performing data mining on these high-dimensional datasets, it usually need execute feature selection. Feature selection can reduce the execution time of the classification algorithms, improve the classification accuracy of the classifier, simplify the classification model and improve the understandability of the classification model by removing irrelevant, redundant and noise features before training the classification model.With the advent of the era of big data, feature selection technology has become a hot topic in the field of machine learning, datamining, and pattern recognition and so on.