Skip to content

An alternative algorithm for the weighting module of PlayeRank. ⚙️ SVM and Logisitic regression implemented by scratch (as it was the task of the project) in order to obtain a vector of parameters that assigns weights to different events in football in order to evaluate a player or a team. 🥅

Notifications You must be signed in to change notification settings

LeonardoAcquaroli/playerank-weighter

Repository files navigation

PlayeRank weighter module from scratch

An alternative algorithm for the weighting module of PlayeRank.
SVM and Logisitic regression algorithms are implemented by scratch (as it was the task of the project) in order to obtain a vector of parameters that assigns weights to different events in football in order to evaluate a player or a team (like a pass, a shot or a foul). Models performance is benchmarked also against non-parametric models like k-NN and Decision tree, always implemented from sratch. 🔢

Repository for the Project of Machine learning at Unimi.

About

An alternative algorithm for the weighting module of PlayeRank. ⚙️ SVM and Logisitic regression implemented by scratch (as it was the task of the project) in order to obtain a vector of parameters that assigns weights to different events in football in order to evaluate a player or a team. 🥅

Topics

Resources

Stars

Watchers

Forks