3.3.5. Root Mean Square Error

The root mean square error (RMSE) is a measure of accuracy for comparing the forecast errors of different models based on the same dataset. It is the square root of the average of the squared errors, mathematically computed as follows

$$RMSE = \sqrt{\frac{\sum\_{i=1}^{n} (P\_i - A\_i)^2}{n}}.\tag{14}$$

Since computing the RMSE involves squaring the difference between the predicted and the target values, thus, a few large differences will definitely increase the RMSE compared to the MAE. Consequently, the RMSE is sensitive to outliers, and hence useful for analyzing models with outlier tendencies.

#### **4. Results and Discussion**

In this section, we present and discuss both quantitative and qualitative results obtained from the evaluation and analysis of the ML models considered in our study. By quantitative analysis, we present and discuss the evaluation metrics as they relate to the performance of the different models. By qualitative analysis, we refer to the visual assessment of the different displays of the predicted against the target values of the different models. To this effect, firstly, we conducted a parameter tuning exercise toward ensuring that all models are evaluated based on their best parameter values. For this purpose, the GridSearchCV tool was used with discrete sets of parameter values designated per model. The system hourly demand dataset was used for the fine-tuning process. Thereafter, the fine-tuned models were tested and compared based on the hourly renewable generation

dataset, and the results are discussed. The models were simulated using the Waikato Environment for Knowledge Analysis platform upon on a computer having an i7-10750H central processing unit and an NVIDIA GeForce GTX 1650 Ti GPU.
