Skip to content
forked from osroca/upordown

Predicting whether a stock will go up or down using BigML and data from Quandl and PsychSignal

License

Notifications You must be signed in to change notification settings

beagribes/upordown

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

26 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Up or Down?

Predicting whether a stock will go up or down using data from Quandl and PsychSignal.

Analyzing the data

  1. Install bigmler:

     pip install bigmler
    
  2. Set up your BIGML_USERNAME and BIGML_API_KEY

     [Setting up BigML authentication](https://bigml.com/developers/quick_start#q_authenticate)
    
  3. Run the script

     ./upordown.py
     [2014-05-01 22:17:05] Creating sources...
     [2014-05-01 22:17:20] Creating datasets...
     [2014-05-01 22:17:39] Merging datasets...
     [2014-05-01 22:18:03] Splitting dataset...
     [2014-05-01 22:18:07] Creating a model using the training dataset...
     [2014-05-01 22:18:10] Evaluating model against the test dataset...
     [2014-05-01 22:18:13] Creating an ensemble using the training dataset...
     [2014-05-01 22:18:30] Evaluating ensemble against the test dataset...
     [2014-05-01 22:18:35] Creating model for the full dataset...
     [2014-05-01 22:18:37] Sharing resources...
     [2014-05-01 22:18:40] https://bigml.com/shared/dataset/zy156fHO5woBGSbTuNeJbkg3htM
     [2014-05-01 22:18:40] https://bigml.com/shared/model/jQNDNqxcSDsfKmQp10cIEXV2Jb7
     [2014-05-01 22:18:40] https://bigml.com/shared/evaluation/oQttuHuCiJ0DW01ai1OsVrsGvww
     [2014-05-01 22:18:40] https://bigml.com/shared/evaluation/3w8Baa5yb5sHliJzHMSZJ2PplW8
    

Visualizing the data

Dataset

Sunburst

Single model evaluation

Ensemble evaluation

About

Predicting whether a stock will go up or down using BigML and data from Quandl and PsychSignal

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 62.0%
  • CSS 21.1%
  • JavaScript 16.9%