Parameter Tuning based on user feedback

Hello people,

I am currently researching if there was already some work done on parameter tuning of algorithms based on user feedback?

The alorithms should be visualizable. Based on simple yes/no answers (to the question, "This visualization is good/better") from the user, the parameters should be tuned and presented to the user iteratively as a visualization.

Would be nice if someone could point me to some papers.

Thank you