How to do regression function estimiation by SVM using RBF Kernel along with parameters tuning to predict a future data ahead of a hourly temprature data?

Hello and good day ,

I have a hourly temprature data , so I need to use it to predict the next 24 hours temprature in the future. I tried to use 'fitrsvm' function >> mdl=fitrsvm(X,Y). and then try to predict the future data by 'predict' function >> Y=predict(mdl,Xt), however I only get a one step predection. on another side, the options of using the training function and prediction are confused for me, especially I am a little bit beginnar using such functions. As a result, I can not get what I excatly model I want. So please I will highly appraciate if you help me in that. What I rellay need is :

1> Training data: SVM regression 'function estimation with RBF kernel or any other kernel if it is better and cross validation. Taking in the account the paramters tuning of the model ex C and Kernel parameters .

2> predict future data of the temprature for 24 hours or more.

3> evaluate the model. ex. mse, mape

I did somthing lik the following:

Mdl= fitrsvm(Xtraining,Ytraining,'rbf');

Y= predict(Mdl2,Xt);

and I am not sure if it is correct but the result of y is only one value , and I can not control the options of the model in apprpriate way in order to be bulit as required above.

Please help! Thanks a lot in advance.