Help understand a gradient derivation for RankNET

10 Views Asked by At

I am reading the RankNet to LambdaMART paper : https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/MSR-TR-2010-82.pdf , where the author makes a particular claim in equation (1). They start from a cost function and then explains it. Then, they say "This gives" .

I've been trying to derive it myself and cannot figure it out. Can someone help me understand how they got equation (1) from the definition if "C"?

Specifically, this part:

enter image description here