所有数据来自网易财经 查看源码点击进入Github
Online model vs. Offline model - accuracy
Online model vs. Offline model - f1-score
Learning Curve - AdaBoost
Learning Curve - GNB
Learning Curve - LDA
Learning Curve - RF
sklearn feature_selection 保留特征数 - 评价指标
minimize: Bayesian Information Criterion (BIC) => n=27
短期预测: 股票价格波动更大 "Knowing Future"投资者的收益与震荡幅度直接关联 中长期预测: 预测准确度相比短期更高 长线准确率甚至可以达到90%
© 2016 SJTU@DataMining - IEEE.
