Significance testing against chance
Hi all,
I'm interested in comparing the performance of one model against chance. I know there is the t-test operator which requires 2 performance vector inputs, presumably to compare to models. Is there a way to test the significance of one model against chance with this operator? Perhaps this would require building a "chance" model for comparison, if so, how may I do this?
Thank you.
I'm interested in comparing the performance of one model against chance. I know there is the t-test operator which requires 2 performance vector inputs, presumably to compare to models. Is there a way to test the significance of one model against chance with this operator? Perhaps this would require building a "chance" model for comparison, if so, how may I do this?
Thank you.
0