Research on Optimization of Random Forest Algorithm Based on Spark

As society has developed, increasing amounts of data have been generated by various industries. The random forest algorithm, as a classification is widely used because its superior performance. However, the algorithm uses simple sampling feature selection method when generating subspaces which cannot distinguish redundant features, thereby affecting accuracy, and resulting in low calculation efficiency stand-alone mode. In response to aforementioned problems, related optimization research was conducted with Spark present paper. This improved performs extraction according calculated importance form subspace. When model, it selects decision trees based on similarity accuracy different decision. Experimental results reveal that compared original proposed paper exhibited higher rate could effectively classify data.