Skip to content

Latest commit

 

History

History
28 lines (18 loc) · 1.51 KB

README.md

File metadata and controls

28 lines (18 loc) · 1.51 KB
Raw

image

Sed-Young-Wa

Sed-Young-Wa (เสร็จยังหว่า) is a service for estimating the end of the issues submitted in the Traffy fondue platform. We built a regression model GradientBoostingRegressor with features from the issues including issue types, submission date, and district. And also uses scraped features: weather and holidays in Thailand. After that, We constructed a model pipeline and stored model variables on MLFlow to retrieve the model in production. The model results achieved a root-mean-squared error (RMSE) of 20.

Diagram

Training phase

Run-time phase

Appendix

  1. You can view the model pipeline process in the notebook link.
  2. A clip of the final report is shown here.

Authors

  • @NonKhuna [Data science]: Create a model and preprocess Data.
  • @Karn [Data engineering]: Model pipeline.
  • @Tantawit [Data engineering]: Scraping Data
  • @Anon [Data visualization]