Effect of Selection of Classification Features C4.5 Algorithm in Student Alcohol Consumption Dataset

Abstract

Alcoholic beverages are psychoactive substances that are addictive. Psychoactive substances are a class of substances that work selectively, especially in the brain, which can cause changes in behavior, emotion, cognition, perception and awareness of one's and others. Police survey results in 2014 showed that users of narcotics and liquor. Most of the group of students, both junior, and senior student, which amounts to 70%, while only 20% of primary school graduates. In the modern era, especially in information technology, the need for information and the latest knowledge is multiplying. One of them is the user information of alcohol among teenagers is more accurate. Data mining is the process for extracting and identifying information useful and relevant knowledge from a variety of big data. In the data mining, there is a classification technique that assesses the data objects to include it in a particular class of several classes available, can be applied in the case - the case in the health sector, for example, in the case of alcohol addiction in adolescents. The algorithm that can be used in the classification is the C4.5 decision tree. The use of the decision tree algorithm to determine the level of alcohol use in teenagers using two methods, namely, the selection of attributes and without attributes.