/LTFS-Data-Science-FinHack-2

Primary LanguageJupyter NotebookApache License 2.0Apache-2.0

LTFS-Data-Science-FinHack-2

Used google colab environment

Dataset split:

  1. In "FinHack_V1_Ada_Boost.ipynb" manually created validation set from train set
  2. In "FinHack_AdaBoost_Simple_Model.ipynb" manually created validation set from train set by removing lastest dates from train set
  3. Manually created test set file that is inline with train file

Approach:

  1. Added holidays
  2. Added share prices of L&T Finance Holdings
  3. Added date time features
  4. Used Adaboost Regressor from sklearn

Note: Share price data is not considered for "FinHack_AdaBoost_Simple_Model.ipynb"

Best Model: "FinHack_AdaBoost_Simple_Model.ipynb" with Public score: 13.460, Private score: 23.0178