• Dr.T.R.Sivapriya
• C.Malarvizhi

Online gaming addiction greatly affects the social and psychological wellbeing of people. Early prediction of problem gambling plays a major role in the diagnosis and treatment of the problem gambler. Machine learning algorithms are found to be efficient predictors in finding the major factors that lead to problem cross gambling. Forecasting problem gambling behavior would mitigate the risk of unhealthy gambling behavior. Missing data and unbalanced real time datasets influence the classification accuracy to mine gambling behaviors. Preprocessing of dataset improves the performance of classifier. Three single Imputation and three hybrid techniques are compared by the accuracy provided by random forest. Class balancing of dataset improved the classification accuracy of random forest. Bayesian networks and random forest classifiers were used for prediction of problem cross gambling. Imputation followed by class balancing yielded an accuracy of 96.5% and AUC nearly 0.98 with random forest classifier. The improved accuracy would enhance early prediction for early intervention and medical care. The proposed categorical hybrid feature selection technique explored the optimal feature subset for the effective intervention in cross gambling that could be applied in similar online games, with an accuracy of 99.73.

Problem gambling, classification, imputation, random forest.

"PREDICTIVE ANALYSIS OF KEY FACTORS THAT INFLUENCE HIGH RISK CROSS GAMBLING WITH CLASS BALANCED CATEGORICAL HYBRID FEATURE SELECTION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 5, page no.35-39, May-2019, Available :http://www.jetir.org/papers/JETIR1905007.pdf

"PREDICTIVE ANALYSIS OF KEY FACTORS THAT INFLUENCE HIGH RISK CROSS GAMBLING WITH CLASS BALANCED CATEGORICAL HYBRID FEATURE SELECTION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 5, page no. pp35-39, May-2019, Available at : http://www.jetir.org/papers/JETIR1905007.pdf