https://github.com/hariprasath-v/machinehack_predict_accident_risk_score_for_unique_postcode

Machine Learning model to predict Accident Risk Index by using area/district level accident data
https://github.com/hariprasath-v/machinehack_predict_accident_risk_score_for_unique_postcode

exploratory-data-analysis jupyter-notebook klib matplotlib pandas pycaret python regression-models seaborn

Last synced: 7 months ago
JSON representation

Machine Learning model to predict Accident Risk Index by using area/district level accident data

• Host: GitHub
• URL: https://github.com/hariprasath-v/machinehack_predict_accident_risk_score_for_unique_postcode
• Owner: hariprasath-v
• License: apache-2.0
• Created: 2022-03-29T05:13:39.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2022-04-06T12:30:12.000Z (over 3 years ago)
• Last Synced: 2025-01-13T01:44:53.972Z (9 months ago)
• Topics: exploratory-data-analysis, jupyter-notebook, klib, matplotlib, pandas, pycaret, python, regression-models, seaborn
• Language: Jupyter Notebook
• Homepage:
• Size: 3.27 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Machinehack_predict_accident_risk_score_for_unique_postcode

### Competition hosted on  MACHINEHACK 

# About

### Machine Learning model to predict Accident Risk Index by using area/district level accident data.

### Competition Public LB Rank:201   &  Private LB Rank:200

### Final Score  0.6497

### Evaluation Metric is Mean squared error.

### File information

 * MH_predict_accident_risk_score_for_unique_postcode_EDA.ipynb

 

    ### Packages Used,

        * seaborn

        * Pandas

        * klib

        * Numpy

        * Matplotlib

        * re

    

    ### Basic Exploratory Data Analysis

    

* MH_predict_accident_risk_score_for_unique_postcode_Model.ipynb

    ### Packages Used,

        * Sklearn

        * re

        * Pandas

        * Numpy

        * Matplotlib

        * pycaret

        * shap

      

     ### Data Pre-processing         

     ### Created  pycaret voting regressor model.

     ### Model interpretation with shap  

     

### Residuals Plot

![Alt text](https://github.com/hariprasath-v/Machinehack_predict_accident_risk_score_for_unique_postcode/blob/main/Pycaret%20Voting%20Regressor%20Residuals%20plot.png)

### Prediction Error Plot

![Alt text](https://github.com/hariprasath-v/Machinehack_predict_accident_risk_score_for_unique_postcode/blob/main/Pycaret%20Voting%20Regressor%20Prediction%20Error%20Plot.png)

### Pycaret Feature Importance Plot

![Alt text](https://github.com/hariprasath-v/Machinehack_predict_accident_risk_score_for_unique_postcode/blob/main/Pycaret%20Catboost%20Regressor%20Feature%20Importance%20Plot.png)

### SHAP Feature Importance  

![Alt text](https://github.com/hariprasath-v/Machinehack_predict_accident_risk_score_for_unique_postcode/blob/main/Shap%20Feature%20Importance%20Plot.png)

### SHAP Feature Impact

![Alt text](https://github.com/hariprasath-v/Machinehack_predict_accident_risk_score_for_unique_postcode/blob/main/Shap%20Feature%20Impact%20on%20Model%20Plot.png)

### SHAP Feature Impact on Single observation

![Alt text](https://github.com/hariprasath-v/Machinehack_predict_accident_risk_score_for_unique_postcode/blob/main/Shap%20Single%20Observation%20Feature%20Impact%20Plot.png)