ComputingComputer ScienceHigh-performance ComputingData MiningMachine LearningParallel ComputingClassificationData ParallelismDecision TreesDistributed ComputingScalable Data MiningScalable Machine LearningAccuracyDistributed Data MiningParallel Data MiningParallel Machine LearningRandom Forest ClassificationRandom ForestsScalabilityScalable Classification
In order to improve the efficiency and adaptability of classical random forest algorithm in large data environment, an improved random forest algorithm based on Spark is proposed. Firstly, an improved random forest algorithm (FRF) based on the Fayyad boundary point principle is proposed to deal with the shortcomings of classical random forest algorithm in the process of discretization of continuous attributes. Then, using Spark's memory computing advantages, an improved random forest parallelization algorithm based on Spark is proposed. The experimental results show that FRF can effectively shorten the construction time while maintaining the computational accuracy, and it has good scalability and high efficiency in the distributed environment, and can perform fast classification task in large data background.
9
Leo Breiman · Machine Learning · 2001 · 126.2K citations · Full text
Sanjay Ghemawat · Communications of the ACM · 2008 · 18.5K citations · Full text
Leo Breiman · Machine Learning · 1996 · 17K citations · Full text