Abstract
Drug abuse is a problem that can be influenced by various interrelated factors. This study aims to apply Mamdani Fuzzy logic in predicting the level of drug abuse in Padang City based on several factors obtained from documentation and interviews. The input variables used consist of the Age of the Perpetrator, Environment, Socializing, Gender, and Family, while the output variable is the Level of Drug Abuse with Low and High categories. The research stages include data collection, determining variables and fuzzy sets, fuzzification, forming if-then rules, inference using the minimum operator, composition using the maximum operator, and defuzzification using the centroid method. A total of 64 fuzzy rules are used to connect the combination of input variables with the output. Testing is done through manual calculations and implementation using MATLAB. The test results show that the output categories from manual calculations and MATLAB are the same in three test data, namely High, High, and Low. These results indicate that the Mamdani Fuzzy method can be used in modeling the level of drug abuse based on the variables used.
