TABLE 8.

Evaluation metrics

Measure	Formula	Remarks
$R^2$	$R^2=1-\left(\sum \left(Y_i-X_i\right)^2/\sum \left(Y_i-Z_i\right)^2\right)$	R‐Square score tells how close the forecasted value is to the regression line (Chordia & Pawar, 2021). The R‐squared (R 2) score is a simple to compute and explain gauge of confidence (Lupón & Gaggin, 2015). It is the level to which a data point fits the linear regression, thereby indicating how well the regression line forecasts real values. The coefficient of determination gives you to compute how much variance the model's self‐standing variables have displayed. It gives us the model's goodness‐of‐fit, as well as a score that is always between 0 and 1 (Renaud & Victoria‐Feser, 2010). Whereas the true value at certain ith assumption, is the forecasted value at ith assumptions, shows the mean of the overall observations, and is the overall count of observations
RMSE	RMSE = $\surd \sum \left(\left(Y_i-X_i\right)^2\right)/n$	RMSE is defined as the variance or root of the residuals. It's a metric for how the residuals distribute around the best‐fit line. It will be easy to read because the units are the same as the output units. This is often negative in nature, and a lower RMSE value enhances model performance. Whereas $Y_i$ shows the predicted value, $X_i$ shows the real value and n is the count of observations. (Moody, 2019; Willmott & Matsuura, 2005)