Author: Tiago Russomanno
Streamlit interface for the final project at Datascientist, focusing on predicting car accidents in France.
The primary goal of this project is to predict the severity of road accidents in France based on historical data. This project encompasses all stages of a Data Science project, providing a comprehensive exploration of predictive modeling. The project workflow includes:
Data Cleaning: Study and application of methods to clean the dataset, ensuring quality input for the predictive model.
Feature Extraction: Extraction of relevant characteristics from historical data to estimate accident severity.
Scoring of Risk Zones: Utilizing model results to score risk zones based on meteorological information, geographical location (GPS coordinates), satellite images, etc.
Model Training: Development of a predictive model using machine learning techniques.
Model Comparison: Comparison of the trained model's predictions with historical data.
Data.gouv.fr - Bases de données annuelles des accidents corporels de la circulation routière (2005-2019)
Kaggle - Accidents in France (2005-2016)
Bases de données annuelles des accidents corporels de la circulation routière - Années de 2005 à 2020 - data.gouv.fr
The validation conditions for the project include:
An exploration, data visualization, and data pre-processing report.
A modeling report.
A final report and associated GitHub repository.
Link to the project document: https://docs.google.com/document/d/1m2ibEY6n6zcnqqxuJWyTgvIjmhyGQADBCmZThVQnQpA/edit