/traffic_forcasting

Primary LanguageJupyter Notebook

traffic_forcasting

This is a template to help beginners work on the traffic forcasting program. I will add more lines to the code. Here are the plans:

  1. Parse the day and page. (done:07/18/2017)
  2. Split train, test and hold out, and build a random forest model. (done:07/19/2017)
  3. Build a function for the metric. (done:07/18/2017,sampe.py)
  4. Build a function for submission.
  5. Add google trend as feature.
  6. Add simple time-series features. (done:07/19/2017)
  7. Build more features using clustering.