LSTM时序预测滞后现象
| | 分类于 LSTM
字数统计:343字 | 阅读时长:1 分钟
阅读量: 0 | 评论量:

LSTM时序预测滞后现象

0 Views LSTM with
本文字数:343 字 | 阅读时长 ≈ 1 min

在使用LSTM进行时间序列预测时,有时会遇到预测滞后问题,也就是LSTM的预测值滞后于真实值的变化。出现滞后现象的一些文章和可能解释收集如下

  1. 用 LSTM 做时间序列预测的一个小例子
  2. 时间序列预测—-预测的结果跟实际的时间序列值存在滞后
  3. 代码干货 | 基于Keras的LSTM多变量时间序列预测
  4. 但如果只用time series数据,你很有可能得到的就是滞后一天的趋势
  5. 一个带输出图像的LSTM预测
  6. 用LSTM预测时间序列存在延迟现象?
  7. Stackexchange problem
  8. Stackoverflow-Delay issue in time series prediction
  9. LSTM for time series prediction
  10. LSTM Neural Network for Time Series Prediction
  11. 知乎问题-Pyhong的答案
  12. https://dashee87.github.io/deep%20learning/python/predicting-cryptocurrency-prices-with-deep-learning/
  13. https://medium.com/@siavash_37715/how-to-predict-bitcoin-and-ethereum-price-with-rnn-lstm-in-keras-a6d8ee8a5109
  14. 最全 LSTM 模型在量化交易中的应用汇总(代码+论文)
  15. https://www.kaggle.com/pablocastilla/predict-stock-prices-with-lstm
  16. useful discussion
  17. another discussion
  18. another another discussion
  19. https://jiasuhui.com/article/3855
  20. 知乎问题
  21. 延时现象解决
  22. github上的讨论
  23. ResearchGate上的讨论
  24. possible solution
    1. randomize training samples in each batch, make sure they are not followed one by one
    2. choose or design a better loss function other than MSE
    3. extract some features from the input time series
    4. manually limit the weight of x_{t-1}, x_{t-2}

ght of x_{t-1}, x_{t-2}

2019-07-09
NLP