I think we should calm down in waiting for their solution, most of them still have to tidy their code and also they have to wait for zindi to review and accept the code before they can share(1st - 3rd), if they want to, also the title of this discussion though...
No i do it manual in a greedy way. I always varied one parameter and fix all the others. For the parameter that I have varied I loop over different values and tested them via cv score. Tuning of the parameters improved my cv score a lot, so i invested a lot of time.
 scale(.21174)'%3e%3cpath d='m52.487 177.086 13.018 11.852 11.762-11.762 11.896 12.266 12.846-11.358 11.827 12.43 13.339-11.013 10.897 12.728 13.27-10.145 13.947 12.319'/%3e%3ccircle cx='89.425' cy='154.584' r='5.167'/%3e%3ccircle cx='111.237' cy='156.047' r='5.167'/%3e%3ccircle cx='30.05' cy='16.5' r='4'/%3e%3c/g%3e%3c/svg%3e)
I think we should calm down in waiting for their solution, most of them still have to tidy their code and also they have to wait for zindi to review and accept the code before they can share(1st - 3rd), if they want to, also the title of this discussion though...
You can found my solution(place 16) in github: https://github.com/berndAllmendinger/Zimnat-Insurance-Recommendation-Challenge
Good
Thank you! Could you please share how you find optimal params? with tools like hyperopt or not?
No i do it manual in a greedy way. I always varied one parameter and fix all the others. For the parameter that I have varied I loop over different values and tested them via cv score. Tuning of the parameters improved my cv score a lot, so i invested a lot of time.
Thank you for the reply. Do you have your rules what to fix first etc?
No i have no rule. More or less random