"Grouping together confidence and prediction results into categories"

GoonerGooner Member Posts: 1 Contributor I
edited June 2019 in Help
Hi all

Sorry guys but I’m new to machine learning so I apologise in advance if I’m explaining things using the wrong terminology.

Basically I’m trying to predict horse racing results and after days of reading and watching tutorials I’ve got as far as building a process with Naïve Bayes algorithm with my data split to apply model and performance to give me a confidence Probability of a win and a win or lose prediction.

but………   my main problem is all the prediction is done on a per horse level and not a per race level, :( so basically I’ve got multiple predicted winners in each race and confidence % that doesn’t match the race.

So for example what i want to achieve is if there were a race with 4 horses all evenly matched (highly unlikely I know lol) then each horse would have a confidence 0.25 or 25% with the total for each race always adding up to 100%

So in a nut shell I want to

1. Predict 1 winner per race
2. Find the probability of each horse per race to win

I’m Really pulling my hair out with this problem so thanks in advance to anyone that can help me! :)


  • MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,453 RM Data Scientist

    what you need is a new attribute with the race date/name. This needs to be set to a special role (id?). This way it is not used in the prediction. Afterwards you can use some of the loops to loop over that. E.g. loop values.

    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
Sign In or Register to comment.