%0 PDF %T Analyzing the Putative Higgs Boson using Support Vector Machines. %A Ghimire, Mukesh. %D 2015-12-07T22:23:05.550Z %8 2015-12-07 %I Tufts University. Tisch Library. %R http://localhost/files/8s45qm86w %X Support vector machines (SVMs) provide the mathematical framework for a powerful data classifier, which are of great interest in experimental physics. In this project I tested the classifying efficacy of SVMs against that of boosted decision trees (BDTs), which was the data classifier used to discover the Higgs boson in 2012. We tested various parameters that affected the performance of the classifiers, and trained them to classify sets of data with two, three, and eight features. SVMs demonstrated better classifying ability across the board, but at a cost of computational power that increased more in proportion to the complexity of the data. In conclusion, SVMs are better than BDTs at classifying data with a small number of features, but for a large number of features, the choice of the better classifier depends on the computational power at hand. Submitted in partial fulfillment of the grant requirement of the Tufts Summer Scholars Program. %G eng %[ 2022-06-10 %~ Tufts Digital Library %W Institution