• Version
• Download 12
• File Size 306.58 KB
• File Count 1
• Create Date 12 June 2025
• Last Updated 12 June 2025

Wildfire Risk Assessment System

Authors:

1st Mr.P. Rajapandiyan 1 ,2nd A.Akash , 

1Associate Professor and Head of Department of computer Applications, Sri Manakula Vinayagar Engineering College

(Autonomous), Puducherry 605008, India

2Post Graduate student, Department of computer Applications, Sri Manakula Vinayagar Engineering College (Autonomous),

Puducherry 605008, India

Akashappar18@gmail.com

*Corresponding author’s email address: akashappar18@gmail.com

Abstract: This project aims to develop a predictive model for forest fire detection using historical fire datasets and weather report features. Leveraging Data Science and Machine Learning techniques, the model learns from past fire incidents—incorporating factors such as temperature, humidity, wind speed, and rainfall—to detect the likelihood of future fires. The system is built as a web application using Flask, offering real-time fire risk predictions through a simple user interface. The project pipeline includes data ingestion and cleaning, exploratory data analysis and visualization, feature engineering, model training with classification algorithms, and model evaluation using accuracy, precision, recall, and ROC curves. The final model is deployed through a user-friendly Flask web application. This system enables stakeholders such as environmental agencies and local authorities to make informed decisions and take timely preventive actions against potential fire outbreaks.

Keywords: Forest Fire Prediction, Machine Learning, Data Science, Flask Web Application, Predictive Analytics, Fire Weather Index, Classification, Regression.