Skip to content

Just proof-of-concept. Retrain and compare models every <?> period, display on frontend.

Notifications You must be signed in to change notification settings

WillHTam/bitcoin-streamer-lstm

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Stack

How to run

  • install dependencies with yarn

  • any changes to the frontend must be accompanied by yarn build in the 'client' folder

  • run with node server.js

  • view frontend on http://localhost:3000

  • processor should be run with cronjob

  • need to sudo chmod 777 processsor to disable read-only on notebooks

  • gcloud compute ssh <INSTANCE NAME>

TODO:

  • Add Firebase DB to save previous values, accuracy, etc.
  • Tune hyperparameters and features for XGBoost
  • Implement Pickling for SKLearn, DMatrix for XGBoost, and model saving for Keras
  • Replace Random Forest with ARIMA?

Production:

  • Host Processor on different server
  • Run weight processing with Go / Java
  • Use HTTPS
  • Use PGP

About

Just proof-of-concept. Retrain and compare models every <?> period, display on frontend.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published