An Efficient Feature Weighting Method for Support Vector Regression

Support vector regression (SVR) is a powerful kernel-based method which has been successfully applied in problems. Regarding the feature-weighted SVR algorithms, its contribution to model output taken into account. However, performance of subject feature weights and time consumption on training. In paper, an efficient proposed. Firstly, value constraint each weight obtained according maximal information coefficient reveals relationship between input output. Then, constrained particle swarm optimization (PSO) algorithm employed optimize hyperparameters simultaneously. Finally, optimal are used modify kernel function. Simulation experiments were conducted four synthetic datasets seven real by using proposed model, classical SVR, some state-of-the-art models. The results show that superior generalization ability within acceptable time.