Advancements in Communication and Systems
Explorative Analysis for Predicting Direct and Indirect Affected Population due to Alcohol Abuse in Karnataka using Machine Learning Techniques
Authors: Neha Dhirendra Sirur, Shreyas S. Airani, Amogh R. Mangalvedi, Nischay N. Sheshadry, P. G. Sunitha Hiremath and Tulasa A. Badagi
Publishing Date: 13-09-2024
ISBN: 978-81-955020-7-3
Abstract
The objective of the work is to predict directly and indirectly affected population in Karnataka, India, due to alcohol consumption by considering five years (2017-2022) of government data on alcohol licensing, sales, consumption, population trends, and treatment centers. Districts are grouped into 7 divisions as recommended by the state excise department. It is observed that the alcohol sales rose by 7.3%, impacting direct (13.6%) and indirect (13.7%) affected populations. The average alcohol consumption of an alcohol dependent person per day is around 239. 5ml.During the period 2020-23 the alcohol consumption increased by 3.7%, signifying high risk. The directly affected population with high-risk is around 53.2lakhs. The existing treatment infrastructure comprises of 427 centers, encompassing private, IRCAs, and outpatient hospitals. Each treatment center, when compared against the directly affected population, attends approximately 18,300 individuals annually. On a monthly scale, this equates to 1,531 individuals per hospital, assuming a modest 20-bed capacity. This leaves a staggering 52.17 lakh people untreated annually, highlighting a substantial gap between those seeking treatment and the available resources. With only 34 IRCAs, our findings stress the need to expand rehabilitation facilities to effectively combat alcohol addiction. We have also employed Machine Learning algorithms i.e. the Linear Regression model to predict directly and indirectly affected population. The model predicted the directly and indirectly affected population with an accuracy of 96% and 94% and RMSEs of 0.0206 and 0.052, respectively. Finally, we substantiate our findings and provide recommendations to expand rehabilitation facilities, also enhancing public awareness on responsible drinking, and advocating policy reforms to address the treatment gap effectively
Keywords
Licenses, Sales, De-addiction center, Directly affected population, Indirectly affected population, Policies
Cite as
Neha Dhirendra Sirur, Shreyas S. Airani, Amogh R. Mangalvedi, Nischay N. Sheshadry, P. G. Sunitha Hiremath and Tulasa A. Badagi, "Explorative Analysis for Predicting Direct and Indirect Affected Population due to Alcohol Abuse in Karnataka using Machine Learning Techniques", In: Ashish Kumar Tripathi and Vivek Shrivastava (eds), Advancements in Communication and Systems, SCRS, India, 2024, pp. 557-566. https://doi.org/10.56155/978-81-955020-7-3-49