PREDICTIVE ANALYTICS OF TOURISM IN INDIA: A DATA MINING APPROACH TO VISITOR FORECASTING

Abstract

The tourism sector contributes extensively to regional and national economic progress; therefore, reliable visitor prediction has become important for planning, infrastructure utilization, and policy formulation. This study examines the use of predictive analytics and machine learning approaches to estimate tourist inflow patterns by analyzing quantitative tourism datasets collected from Kaggle and global tourism databases.

The study utilizes historical tourism data, including worldwide tourist arrivals, travel receipts, GDP, inflation, unemployment, and seasonal trends from 1999 to 2023. Prior to building the predictive models, the gathered dataset was processed through several preprocessing techniques such as data normalization, handling of missing entries, and feature refinement.

Three machine learning algorithms—Decision Tree, Random Forest, and Support Vector Machine (SVM)—were applied and evaluated using performance measures including Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Accuracy Score.

The quantitative findings showed that the Random Forest algorithm delivered the best forecasting performance with an accuracy rate of 93.4%, exceeding the results obtained from the Decision Tree and SVM models. The outcomes reveal that advanced machine learning and ensemble-based techniques can substantially enhance tourism prediction by identifying hidden and complex visitor behavior patterns.

The study contributes to smart tourism research by providing a scalable predictive framework that supports tourism policy planning, resource allocation, and destination management.