Tools
Tools
Vasamsetti, Ram Abhilash (2023) Using Time Series Predictive Models for Early Detection of Gambling Addiction in Problem Gamblers. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (1MB) | Preview |
Abstract
Gambling is a game of chance and addictive in nature due to its uncertain lucrative outcomes. Gambling Addiction has become one of the most challenging aspect for gambling operators and governing bodies as it negatively impacts the user and their immediate family. Early Detection and Prevention becomes complicated due to subtle behavioural changes and patterns. Modern technology have opened doors to multiple platforms available both offline and online that allow users to engage in gambling on almost any events of change such as sports outcomes to simple lottery games.
This work focuses on early detection of gambling addiction in players who engage in online betting by analysing their bet wagering patterns and other demographic data. This paper combines K means clustering algorithm with LSTM and ARIMA/SARIMA time series forecasting models to identify various groups of players based on their demographic and wagering activity and forecast their future betting patterns. High risk players’ future betting patterns are forecasted using predictive models. Identifying high risk betting early will help in providing Responsible Gambling interventions to safeguard players mental health and limit financial losses.
The work clusters the players based on their Demographic and Wagering data into 3 groups. The players with problem gambling are successfully identified by analysing the mean of features. The wagering patterns in this cluster are forecast and validated using RMSE and MAE. An Average RMSE of 1247.9 is obtained for ARIMA model and better 1239.42 RMSE for SARIMA model. LSTM received a RMSE of 1524.24 despite solving highly non stationary data points. The forecast graphs are plotted for each user showing their predicted bet wagering pattern.
The 2 model architecture has gained new insights in the domain of Gambling addiction detection and prevention where not only high risk gamblers are identified but also possible future patterns where a player could engage in high bet wagering, are predicted. The findings of this paper will enable a foundation of a method to closely monitor high risk and potential risk players and provide early interventions to prevent them from gambling addiction risks.
| Item Type: | Thesis (Masters) |
|---|---|
| Supervisors: | Name Email Staikopoulos, Athanasios UNSPECIFIED |
| Subjects: | Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science G Geography. Anthropology. Recreation > GV Recreation Leisure > Games and Amusements > Gambling Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning |
| Divisions: | School of Computing > Master of Science in Data Analytics |
| Depositing User: | Ciara O'Brien |
| Date Deposited: | 23 May 2025 14:54 |
| Last Modified: | 23 May 2025 14:54 |
| URI: | https://norma.ncirl.ie/id/eprint/7630 |
Actions (login required)
 |
View Item |