Parameters Optimization Improvement of SVM on Load Forecasting
暂无分享,去创建一个
The SVM regression algorithm has shorter convergence time, high precision, and less adjustable parameters, the structure can be confirmed difficultly in the power load forecasting. In view of the grid search method of parameter optimization in support vector machine (SVM) exists the shortage of high complexity and computation, an improved grid search method is revealed, which is based on the idea of uniform particles, this paper apply it to the load forecast. Compared with the traditional grid method, the improved grid search method can effectively reduce the computational complexity and ensure the accuracy of the prediction precision. We illuminate the method of new ideas, and prove the feasibility of this view through the experiment.
[1] Ming-guang Zhang,et al. Parameter optimized LS-SVM based on APSO for short-term load forecasting , 2014 .
[2] Habibollah Haron,et al. Statistical Learning Theory and Support Vector Machines , 2010, 2010 Second International Conference on Computer Research and Development.
[3] Yanyu Chen,et al. Modified Parallel Cat Swarm Optimization in SVM Modeling for Short-term Cooling Load Forecasting , 2014, J. Softw..