A modified projective forward-backward splitting algorithm for variational inclusion problems to predict Parkinson's disease

Abstract

This research studies variational inclusion problems, which is a branch of optimization. A modified projective forward-backward splitting algorithm is constructed to solve this problem. The algorithm adds the inertial technique for speeding up the convergence, and the projective method for several regularization machine learning models to meet good model fitting. To evaluate the performance of the classification models employed in this research, four evaluation metrics are computed: accuracy, F1-score, recall, and precision. The highest performance value of 92.86% accuracy, 62.50% precision, 100% recall, and 76.92% F1-score shows that our algorithm performs better than the other machine learning models.

Keywords

• Variational inclusion problem
• extreme learning machines (ELM)
• forward-backward splitting algorithm
• regularized least squares
• Parkinson's disease
• 46T99