Hi, I have a problem set to predict winners of Running races. The data I got was groups of data, in each group I have the attributes of each runner and their best time ranking in the corresponding group. For each group, the number of runners is different. So there may be 2 runners in a group and the slowest one rank 2 but the time he achieved could be ranking 10th if he is in a group of 12 people. But I don't have the exact timing of each runner. As such, which machine learning model would be useful that could be trained based on a group of runner, instead of being trained by each individual runner?
I have tried the supervised learning models, like logistic regression, KNN, decision tree, SVM, etc. But the score I got was around ~75% only. Would there be way to increase the accuracy?
Top comments (0)