regression learner svm load forecast

Hello I'd like to use regression learner in predicting the load of a residential unit using historical data and temperature records as inputs. I'll be using SVM as my method, and I'm not sure of the right way to represent the inputs. I have the vectors of the historical data and the temperatures and I can import them, but how do I set it up so that for example; use the historical load of the last 10 hours to detect the the next hour and so on. I mean which data is imported as a response and which is imported as predictor.

Thanks