An Epsilon Hierarchical Fuzzy Twin Support Vector Regression
This is an incremental improvement for regression tasks in machine learning, specifically addressing uncertainty in forecasting.
The paper tackled forecasting problems with uncertainty by developing epsilon hierarchical fuzzy twin support vector regression (epsilon HFTSVR), which improved regression performance and achieved remarkable generalization with minimum training time on synthetic and real datasets.
The research presents epsilon hierarchical fuzzy twin support vector regression based on epsilon fuzzy twin support vector regression and epsilon twin support vector regression. Epsilon FTSVR is achieved by incorporating trapezoidal fuzzy numbers to epsilon TSVR which takes care of uncertainty existing in forecasting problems. Epsilon FTSVR determines a pair of epsilon insensitive proximal functions by solving two related quadratic programming problems. The structural risk minimization principle is implemented by introducing regularization term in primal problems of epsilon FTSVR. This yields dual stable positive definite problems which improves regression performance. Epsilon FTSVR is then reformulated as epsilon HFTSVR consisting of a set of hierarchical layers each containing epsilon FTSVR. Experimental results on both synthetic and real datasets reveal that epsilon HFTSVR has remarkable generalization performance with minimum training time.