Detection Of Fraudulent Behaviour In Drinking Water Consumption Using KNN And SVM Algorithms.

This project focuses on detecting Detection Of Fraudulent Behaviour In Drinking Water Consumption using machine learning techniques. The dataset used contains various features, such as "Avg_Monthly_Usage_KL," "Current_Month_Usage_KL," "Billing_Amount," "Payment_Delay_Days," and "Previous_Fraud_Flags," along with a target variable, "Fraud_Label," which indicates whether an account is labeled as fraudulent (1) or non-fraudulent (0). The project applies machine learning algorithms like Support Vector Machine (SVM) and k-Nearest Neighbors (KNN) for classification tasks. These algorithms analyze usage patterns, payment behaviors, and historical fraud information to classify accounts into fraud or non-fraud categories. A user-friendly web application is developed using Flask, HTML, CSS, and JavaScript, allowing users to log in, register, and input their water usage data to detect potential fraud. This solution enhances the efficiency of identifying fraudulent activities in water systems, aiding in better resource management and fraud prevention.

Keywords: Fraud Detection, Water Systems, Fraud Detection, SVM, KNN, Flask, Machine Learning, Fraud Classification, Predictive Maintenance, Water Usage Analysis.