A Novel Approach for Building Cyber Crime Prediction and Analysis Model using Random Forest

According to Indian Cyber Crime Coordination Centre, In May 2024, Around 7,000 cybercrime complaints were registered per day, a jump of 113.7% between 2021–2023 and 60.9% from 2022–2023, and 85% of them were financial online frauds. Cybercrimes can result in significant financial losses, compromise sensitive data, disrupt critical infrastructure, and erode public trust in digital systems. There is a need for analyzing the trends and patterns of cyber crimes in different parts of India. This paper focuses on building a model for analyzing and predicting cyber crimes rate in different parts of India in future. The datasets from NCRB (National Crime Record Bureau), the official Indian government website, are used for data collection. Random forest technique is applied on preprocessing data to build the predictive model and XGBoost is used on huge data for accurate prediction and efficient decision-making. High prediction accuracy for cybercrimes is demonstrated by the evaluation results, providing a solid means of bolstering cyber secure defenses. An 80:20 ratio was chosen to divide the data into training and testing sets since it was shown to be ideal for model accuracy. A number of machine learning models were examined, Random Forest Regressor was selected due to its accuracy and robustness. Metrics for evaluating the model, such as Mean Squared Error (MSE) and R-squared, showed how well the model predicted trends in cybercrime in the future. The HTML, CSS, and JavaScript were used for the good performance of the application's front end to deliver a user-friendly interface for data display and interaction, making the results accessible and useful for stakeholders.