Flood Sense: Machine Learning-Based Flood Risk Prediction and Optimal Land Selection Using Environmental Data Analysis

Abstract

Floods are among the most devastating natural disasters, causing significant damage to life and property. This paper presents an advanced flood prediction model using machine learning techniques to assess flood risks and suggest optimal land selection for development. The system integrates Random Forest for structured environmental data analysis and a Convolutional Neural Network using VGG16 for satellite image classification. A web-based interface was developed to make predictions accessible to users. The combination of numerical and image-based flood assessment provides a comprehensive and reliable method for proactive disaster management and urban planning.