#调整随机森林的参数(调整n_estimators随机森林中树的数量默认10个树,精度递增显著) from sklearn import datasets X, y = datasets.make_classification(n_samples=10000,n_features=20,n_informative=15,flip_y=.5, weights=[.2, .8]) import numpy as np training = np.random.choice([True, False]
tree based ensemble algorithms 主要介绍以下几种ensemble的分类器(tree based algorithms) xgboost lightGBM: 基于决策树算法的分布式梯度提升框架 GBDT(Gradient Boosting Decison Tree) 随机森林 Why is it called random forest 决策树 tree based ensemble algorithms 原始的Boost算法是在算法开始的时候,为每个样本赋上一个权重