Hamburg 2016 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 12: Experimentelle Methoden I

T 12.4: Talk

Monday, February 29, 2016, 11:45–12:00, VMP8 SR 105

Performance and optimization of support vector machines in high-energy physics classification problems — •Mehmet Özgür Sahin, Dirk Krücker, and Isabell Melzer-Pellmann — DESY, Hamburg, Germany

In this talk, the use of Support Vector Machines (SVM) is promoted for new-physics searches in high-energy physics. We developed an interface, called SVM HEP Interface (SVM-HINT), for a popular SVM library, LibSVM, and introduced a statistical-significance based hyper-parameter optimization algorithm for the new-physics searches. As example case study, a search for Supersymmetry at the Large Hadron Collider is given to demonstrate the capabilities of SVM using SVM-HINT.