I want to forecast a time series using the support vector Regression (SVR) model. After writing the code related to the SVR equations, I used the "NMaximize" function to obtain the coefficients of this model, but the input matrix has dimensions of {187000,4} , which running of this function takes so long that I have not been able to get the output, while for dimension of {176,4} I get the output answer in about 2 minutes The following code is for dimension of {176,4} :
To reduce the computation time, I reduced the "MaxIterations" to 10, which also reduced the accuracy. Can anyone suggest me a solution to this problem?