README.md

Machinehack-analytics-olympiad-2022

Competition hosted on Machinehack

About

Create a machine learning model to help an insurance company understand which claims are worth rejecting and the claims that should be accepted for reimbursement.

The Final Competition score is 0.68081

Leaderboard Rank is 24

The Evaluation Metric is Logloss.

File information

• machinehack-analytics-olympiad-2022-eda.ipynb

  Basic Exploratory Data Analysis

  Packages Used,

   * seaborn
   * Pandas
   * Numpy
   * Matplotlib
  

• machinehack-analytics-olympiad-2022-model.ipynb

  Data Pre-processing and model.

  Packages Used,

    * Sklearn
    * Pandas
    * Numpy
    * Matplotlib
    * catboost
    * optuna
    * shap
  

  Created catboost classifier model and tuned the hyperparameters by using optuna framework. Model evaluated with Logloss.

  For more detailed information about the model.

Catboost model Optimization History - Explains the best score at each trials.

Catboost – SHAP feature importance

Catboost – SHAP top feature impact

Top feature influences for class 1

Top feature influences for class 0

Overall Train and Validation Logloss