Machinehack-dare_in_reality_hackathon

Competition hosted on MACHINEHACK

About

Build a machine learning model that predicts the Envision Racing drivers’ lap times.

Competition Public LB Rank: 97/361 & Private LB Rank: 162/352

Evaluation Metric is RMSLE.

File information

• dare_in_reality_hackathon_2021_model.ipynb

  Packages Used,

   * Sklearn
   * catboost
   * Pandas
   * re
   * klib
   * datetime
   * Numpy
   * Matplotlib
   * gradio
   * Optuna
   * shap
  

  Basic Exploratory Data Analysis

  Created Catboost regressor model and tune the hyperparameters with the optuna framework.

  Model interpretation with shap

  Created Demo Web-App using Gradio library

  Model RMSLE is: 0.22119371161605259

Optuna Optimization History

Feature Importance

Hyperparameter Importance

Demo App Screenshot.