Rough ν-support vector regression

摘要

After combining the classical ν-SVR with the rough theory, we propose a rough ν-SVR. Double εs are utilized to construct the rough margin for rough ν-SVR instead of single ε for the classical ν-SVR, and this rough margin consisting of positive region, boundary region, and negative region yields the feasible set of the rough ν-SVR larger than that of the classical ν-SVR, which makes the objective function of the rough ν-SVR not more than that of the classical ν-SVR. This may lead to the improvement of the performance. Meantime, experimental results on benchmark data sets confirm the validation and feasibility of our proposed rough ν-SVR.