EL之GB(GBR)：利用GBR对回归问题(实数值评分预测)建模

EL之GB(GBR)：利用GBR对回归问题(实数值评分预测)建模

---

输出结果

设计思路

核心代码

wineGBMModel = ensemble.GradientBoostingRegressor(n_estimators=nEst,
                                                  max_depth=depth,
                                                  learning_rate=learnRate,
                                                  subsample = subSamp,
                                                  loss='ls')
wineGBMModel.fit(xTrain, yTrain)

msError = []
predictions = wineGBMModel.staged_predict(xTest)