neural network training for price and load forecasting

For doing Electrical load and price forecadting ,

1.can i prefer NN fit tool ?

2.How do i evaluate to know if the training done is correct ?

3.What are the accpeted/ preferred values of error, reggression R, mse values?

4.Is it ok to do normalisation of the data before training to get preferred results?

5.Is there any particular algorithm to be followed for the doing training ?

[Rik]

Have a read here and here. It will greatly improve your chances of getting an answer.

Answer 1

[Srivardhan Gadila]

To get started refer to the documentation of the following: Function Approximation and Clustering, Deep Learning Tips and Tricks & Define Shallow Neural Network Architectures. Also the examples of Deep Learning Toolbox.

As per my knwoledge, In general there wouldn't be any specific or exact answers to the above questions.

1: Yes, you can try the nftool

2 & 3: You can refer to Analyze Shallow Neural Network Performance After Training

4: Yes normalization is preferred in general. Again it may or may not improve the network performance.

5: No there isn't any particular algorithm you have to follow, refer to Function Approximation and Nonlinear Regression