Baseline energy modeling using artificial neural network - cross validation technique / Wan Nazirah Wan Md Adnan, Nofri Yenita Dahlan and Ismail Musirin

This paper presents a baseline energy model
development using artificial neural networks (ANN) with CrossValidation (CV) technique for a small dataset. The CV technique
is used to examine generalization abilities and model reliability of
a small data. This CV-ANN model is simulated with thirty
different structures using two CV techniques, Random Sampling
Cross Validation (RSCV) and K-Fold Cross Validation (KFCV).
Working days, class days and Cooling Degree Days (CDD) are
used as ANN input meanwhile the ANN output is monthly
electricity consumption. The coefficient of correlation (R) is used
as performance function to check the model accuracy. The results
are compared and best CV-ANN structure with the highest value
of R is selected to develop the baseline energy model. The
comparison reveals that most of the average R values are above
0.8 and it shows that the CV-ANN is capable to train the network
even with small set of data. ANN-KFCV model with 6 neurons in
hidden layer is chosen as the best model with average R is 0.91.