BIPOLAR DISORDER CLASSIFICATION USING MACHINE LEARNING
Keywords:
Bipolar disorder, Machine learning, Classification, Feature selection, Staking, Naive bayes, Multilayer Perceptron, Predictive modeling.Abstract
Bipolar illness presents a significant diagnostic challenge due to its clinical variability and comorbidity with other diseases. Our research addresses these issues by developing a machine learning system aimed at enhancing diagnostic accuracy in clinical settings. We assessed many algorithms, such as J48, Random Forest, SMO, Naive Bayes, Logistic Re- gression, Simple Logistic, and the deep learning model Multilayer Perceptron (MLP), us- ing feature selection and stacking methodologies. With 90.00% accuracy for Naive Bayes and 91.67% accuracy for Stacking, we discovered notable improvements. These results underscore the significance of state-of-the-art machine learning methods for improving the classification of bipolar disorder and delivering more precise diagnostic and treatment tools.
