I am working on an evaluation of algorithms from a specific family of feature-selection algorithms. Using the Bayesian hierarchical correlated t-test, I evaluated each pair of algorithms to obtain the probabilities p(alg1 is better), p(alg1 and alg2 are equivalent) and p(alg2 is better) for each unique pair of algorithms from my set (around 10 different algorithms).

My question is, how can I use these probabilities to infer the overall best algorithm (ideally assign probabilities)?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.