microsoft / qlib

Qlib is an AI-oriented quantitative investment platform that aims to realize the potential, empower research, and create value using AI technologies in quantitative investment, from exploring ideas to implementing productions. Qlib supports diverse machine learning modeling paradigms. including supervised learning, market dynamics modeling, and RL.
https://qlib.readthedocs.io/en/latest/
MIT License
15.26k stars 2.61k forks source link

可以用少量股票的历史数据训练模型吗?(例如训练LGBModel) #1785

Open moesakura opened 4 months ago

moesakura commented 4 months ago

❓ Questions and Help

我发现如果instruments中只填几只股票,例如随便选择的["SH600006", "SH600007"]

Training until validation scores don't improve for 50 rounds [20] train's l2: 0.497512 valid's l2: 0.5 [40] train's l2: 0.497512 valid's l2: 0.5 Early stopping, best iteration is: [1] train's l2: 0.497512 valid's l2: 0.5

然后用这个模型预测的话 pred_scores = model.predict(dataset)

datetime   instrument    score
2018-12-03  SH600006      0.0
            SH600007      0.0

我也使用了optuna 进行搜参,但是一直没有找到可用的参数组合

是我的使用方式有问题吗?

jimrok commented 4 months ago

看这个预测结果,模型没有学到东西,输出的分数都是一样的。也可能是你数据没有很好的做特征处理,机器学习比较依赖特征工程。