Subscribe to DSC Newsletter

Hi all. 
Reading an article about ELO rank I have a question. The probability of the "A" team win is a sigmoid function like 1 / (1 + exp( RankB -RankA))
and after the game we need to update these ranks like Rank_new = Rank_old +- K*(1(0) - probability)  

So the main question is how I can use for example NN( or other algo) for finding "K" parameter to make binary crossentropy minimum. And I hope it musn't be constant (I want to  find dependent from the initial player rating) 

The main my problem that I can't understand is that we need after updating parameters use a new input rank for calculating probability. So every epoch we need to update input 


Views: 376

Reply to This

Replies to This Discussion

Hi

https://lethain.com/genetic-programming-a-novel-failure/

I think your best bet is to use Genetic Algorithms (GA) if you are trying to find the optimum value of K;the link above should take you to a sample GA code provided by Will Larson.

Good luck

RSS

Videos

  • Add Videos
  • View All

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service