Recently, the ability to use multimodal multi-features machine learning (ML) applications has started to revolutionize biomedical research enabling to classify and predict diseases, as well as a better understanding of development and treatment outcomes, outperforming more classical analyses such as regressions6–8. Significant ML Support Vector Machine (SVM) classifier models were found for complex disorders, including schizophrenia, bipolar disorder, and depression, identifying discriminating features9. Our own study8 and others7, have shown that the accuracy of ML models increases by using multimodal, multi-features approaches to describe complex disorders, permitting a variety of measurement domains that could be brought to bear on different aspects of disease pathology7. Indeed, ML studies calculating AUD classifiers/predictive models have employed genetic loci8, psychosocial7, family history8, and electrophysiological (EEG) measurements8 as features in a multimodal analysis. In the current ML study, we have utilized EEG, genetics, medication intake, and demographic as predisposition characteristics to predict AUD remission. EEG measurements, especially resting-state functional connectivity (EEG-FC) have been shown to be a reliable diagnostic tool and classifier in AUD and other brain disorders such as post-traumatic stress disorder, and bipolar
