Schizophrenia Diagnosis and Prediction with Machine Learning Models

Abstract

Schizophrenia is a multifaceted mental disorder with varying levels of proneness, making accurate classification essential for early intervention and effective management. This study investigates the application of machine learning techniques to classify schizophrenia proneness levels based on behavioral and demographic features, including age, fatigue, slowing, pain, hygiene, and movement. Using a dataset of 1,000 samples categorized into five levels of proneness—Elevated, High, Moderate, Low, and Very High—we evaluated the performance of Logistic Regression, Support Vector Machine (SVM), Gradient Boosting, and Decision Tree classifiers. Among the models, Logistic Regression achieved the highest accuracy of 94.2%, demonstrating its effectiveness in capturing feature relationships and its suitability for datasets with linear or near-linear patterns. SVM is closely followed with an accuracy of 93%, showcasing its robustness in handling high-dimensional data and non-linear relationships. Gradient Boosting achieved an accuracy of 88.1%, indicating its ability to model complex patterns through iterative corrections, while the Decision Tree, with an accuracy of 75.6%, served as a baseline, reflecting the limitations of single-tree models for complex datasets.