Investigation of The Increase in Drug Use in Medan City Using The Support Vector Machine (SVM) Method

Abstract

Medan city is currently experiencing a troubling rise in the prevalence of drug abuse, necessitating effective strategies for detection and intervention. This research aims to improve the accuracy of identifying drug users in Medan using the Support Vector Machine (SVM) method. Data for the study were sourced from reputable institutions including the National Narcotics Agency (BNN), North Sumatra Regional Police (Polda Sumut), and the Health Office of Medan City. SVM was employed to analyze these datasets and distinguish between drug users and non-users. The study revealed that SVM achieved an impressive detection accuracy of 98.0%, a notable improvement compared to earlier approaches like Convolutional Neural Networks (CNN), which attained 83.33% accuracy.These findings highlight SVM's effectiveness as a robust tool for accurately identifying drug users. The outcomes of this study are anticipated to aid government entities in crafting targeted policies and strategies to combat drug abuse in Medan. By harnessing SVM technology, law enforcement and healthcare authorities can bolster their capabilities in swiftly and precisely detecting and responding to drug-related issues. This research contributes significantly to advancing methodologies in drug abuse detection, emphasizing SVM's pivotal role in achieving superior detection rates. In conclusion, the application of SVM in this study not only enhances detection accuracy but also underscores its potential as a reliable technology for addressing the growing challenge of drug abuse in urban settings like Medan. Future research could further refine SVM models and explore additional datasets to validate its efficacy in real-world scenarios, thereby strengthening efforts to mitigate the societal impact of drug misuse.