Alcoholism is characterized by an inability to control drinking because of physical and psychological dependence on alcohol. The liver, immune system, brain, heart and other organs of the human body are severely damaged by alcoholism. There is a dearth of reliable standard test methods for alcoholism detection. Alcoholism can be diagnosed using EEG signals, which are recorded by monitoring cerebral cortex brain changes. In this paper, the classification of control and alcoholic signals using discrete wavelet transform (DWT) and CatBoost classifier is proposed. The Raw EEG signals are preprocessed and decomposed into six levels of sub-bands using DWT. The suggested technique was validated using the K-fold cross validation approach with the CatBoost classifier using the statistical and non-linear characteristics that were retrieved from chosen sub-bands. The proposed approach is experimented on UCI Alcoholic EEG database and achieved an accuracy of 99.08%.