Friday
Hi,
In AI studio, after training a regression algorithm, there is a prediction histogram displayed. What is the X-axis on this figure? It states "Y_pred - Y_actual", but predicted result minus actual does not seem to be correct, as the numbers are always between -1 and 1.
Is it "(Y_pred - Y_actual) / Y_actual" ? Or maybe (Y_pred - Y_actual) / Y_max" ?
Thanks for the help.
Sunday
The X-axis of the prediction histogram in AI Studio likely represents normalized residuals. The label "Y_pred - Y_actual" suggests it's the difference between predicted and actual values, but the values being between -1 and 1 indicates normalization:
Normalized Residuals: It could be (Y_pred - Y_actual) / Y_max, where Y_max is the maximum actual value, keeping the range within -1 to 1.
Relative Errors: It might also be (Y_pred - Y_actual) / Y_actual, scaled to fit between -1 and 1.
I hope this will work for you.....