Multi-model Fusion and Optimization Algorithm for Data Classification
Abstract
With the missing value data in multi-feature classification problem, single classification algorithm is not very good to adapt to a variety of data sets, which cause the classification accuracy is not high. In this paper, different machine learning techniques such as decision tree, neural network, gradient boosting machine and bagging were compared and fused to increase the accuracy to a certain extent. In order to better adapt to a variety of datasets and diversity of characteristics, the theory of pattern mixed model and random forest were performed into the classification fusion model, and weighted iterative was calculated. The precision, recall rate, and F-measure prove that of the multi-model fusion algorithm can well address the problem of low precision of single classification algorithm, and obtain better stability, accuracy and generalization ability.