Skip to content

julyisyours/A-Cross-correlation-and--RNN-approach-for-securities-prediction

 
 

Repository files navigation

A-Cross-correlation-and-machine-learning-approach-for-securities-prediction

The research is still undergoing, will be releasing soon.
Some Ideas:

  • when making a prediction, one should perceive the world as a whole.
  • information flows (or pulse) always take time to transport, and the time could be measured by corss-correlation.  
  • the exist of auto correlation would corrupt the cross correlation information and thus need to be preprocessed.
  • Recurrent neural network or SVM is good at capturing hidden relation without making much assumptions.

About

Even the smallest system is infinitely interconnected, which is a sad story for predictors :p

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 93.0%
  • Jupyter Notebook 7.0%