梯度推进分类器保存模型

Question

我训练GradientBoostingClassifier并像这样保存

pickle.dump(clf,open("C:/Users/R2D2/Documents/Python/Supervised/TrainedModel.dat",'wb'))

然后像这样加载保存的模型

loaded_model = pickle.load(open("C:/Users/R2D2/Documents/Python/Supervised/TrainedModel.dat",'rb'))

测试数据的准确率为%94，训练数据的准确率仅为%50，但训练后，训练和测试数据的准确率均为%94。保存后如何更改

This look similar but doesn't help