Forest fire prediction using K - Nearest Neighbour Model

¹ Dept. of AIML, Chalapathi College of Engineering & Technology (A), Guntur, A.P., India
² Dept. of IT, Sreenidhi Institute of Science & Technology (A), Hyderabad, India
³ Dept. of ME, Sreenidhi Institute of Science & Technology (A), Hyderabad, India
⁴ Dept. of CSE, Samskruthi College of Engineering and Technology (A), Hyderabad, India
⁵ Dept. of AIML, Samskruthi College of Engineering and Technology (A), Hyderabad, India

Abstract

Forest fire most dangerous threat of wildfires, this learning, explores the efficiency of the K NN technique for forest fire prediction. With different types of datasets encircling meteorological data, our ML model includes crucial environmental variables to different patterns leading to fire accidents. The KNN model’s nonparametric and spatially aware characteristics make it a compelling choice for capturing local dependencies within the dataset. The KNN algorithm’s simplicity and flexibility make it convenient for handling diverse datasets and redesigning to changing environmental conditions. However, it’s pivotal to acknowledge the importance of feature selection and data preprocessing in enriching the model’s performance. Additionally, continuous monitoring and updating of the model with real-time data are essential for ensuring its reliability in predicting forest fire occurrences.