Random Decision DAG: An Entropy Based Compression Approach for Random Forest

被引：3

作者：

Liu, Xin ^{[1
]}

Liu, Xiao ^{[1
]}

Lai, Yongxuan ^{[2
]}

Yang, Fan ^{[1
,2
]}

Zeng, Yifeng ^{[3
]}

机构：

[1] Xiamen Univ, Dept Automat, Xiamen, Fujian, Peoples R China

[2] Xiamen Univ, Software Sch, Shenzhen Res Inst, Xiamen, Fujian, Peoples R China

[3] Teesside Univ, Sch Comp, Middlesbrough, Cleveland, England

来源：

DATABASE SYSTEMS FOR ADVANCED APPLICATIONS | 2019年 / 11448卷

关键词：

Random Forest; Pre-pruning; Directed Acyclic Graph;

D O I：

10.1007/978-3-030-18590-9_37

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Tree ensembles, such as Random Forest (RF), are popular methods in machine learning because of their efficiency and superior performance. However, they always grow big trees and large forests, which limits their use in many memory constrained applications. In this paper, we propose Random decision Directed Acyclic Graph (RDAG), which employs an entropy-based pre-pruning and node merging strategy to reduce the number of nodes in random forest. Empirical results show that the resulting model, which is a DAG, dramatically reduces the model size while achieving competitive classification performance when compared to RF.

引用

页码：319 / 323

页数：5