arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2109.08010cs.LGstat.ML
收藏

WildWood:一种新的随机森林算法

WildWood: a new Random Forest algorithm

Université Paris Cité(巴黎西岱大学);École normale supérieure(巴黎高等师范学院)

Stéphane Gaïffas, Ibrahim Merad, Yiyang Yu

AI总结:

本研究提出新型随机森林类集成算法WildWood,通过袋外样本指数权重聚合全树子树预测、结合直方图加速分裂查找,在速度和性能上优于标准随机森林等成熟集成方法。

摘要:

我们提出了WildWood(WW),这是一种新型的随机森林(RF)类监督学习集成算法。标准RF算法使用自助法袋外样本计算袋外得分,而WW利用这些样本生成更优的预测结果——其预测由森林中每棵完全生长树的所有可能子树的预测聚合得到。该聚合通过在袋外样本上计算的指数权重实现,借助名为context tree weighting的算法,这些权重可被精确且极高效率地计算。这一改进结合用于加速分裂点查找的直方图策略,使得WW相比标准RF、极端梯度提升算法等成熟集成方法,速度更快且性能更具竞争力。

英文摘要:

We introduce WildWood (WW), a new ensemble algorithm for supervised learning of Random Forest (RF) type. While standard RF algorithms use bootstrap out-of-bag samples to compute out-of-bag scores, WW uses these samples to produce improved predictions given by an aggregation of the predictions of all possible subtrees of each fully grown tree in the forest. This is achieved by aggregation with exponential weights computed over out-of-bag samples, that are computed exactly and very efficiently thanks to an algorithm called context tree weighting. This improvement, combined with a histogram strategy to accelerate split finding, makes WW fast and competitive compared with other well-established ensemble methods, such as standard RF and extreme gradient boosting algorithms.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑