Comparative analysis of machine learning algorithms for schizophrenia detection

As mental and neurological disorders continue to rise globally, research utilizing artificial intelligence to analyse and classify differences in EEG signals is growing rapidly. This study utilises six different machine learning algorithms for detecting schizophrenia (SZ) using multichannel EEG signals. In the initial phase of this study, pre-processing is carried out, followed by the application of 13 distinct feature extraction techniques. The extracted features are subsequently classified using various machine learning algorithms, leading to classification accuracies up to 1.00 in four algorithms which are Decision Tree, Random Forest, Support Vector Machines (SVM) and Gradient Boosting. In addition, 5-fold cross-validation is applied to increase the reliability of the study. The findings indicate that the study achieved remarkable success and demonstrates the potential for effectively detecting schizophrenia using EEG signals.