Weighted Classification Cascades for Optimizing Discovery Significance in the HiggsML Challenge

Lester Mackey, Jordan Bryan, Man Yue Mo
Proceedings of the NIPS 2014 Workshop on High-energy Physics and Machine Learning, PMLR 42:129-134, 2015.

Abstract

We introduce a minorization-maximization approach to optimizing common measures of discovery significance in high energy physics. The approach alternates between solving a weighted binary classification problem and updating class weights in a simple, closed-form manner. Moreover, an argument based on convex duality shows that an improvement in weighted classification error on any round yields a commensurate improvement in discovery significance. We complement our derivation with experimental results from the 2014 Higgs boson machine learning challenge.

Cite this Paper

BibTeX
@InProceedings{pmlr-v42-mack14, title = {Weighted Classification Cascades for Optimizing Discovery Significance in the HiggsML Challenge}, author = {Mackey, Lester and Bryan, Jordan and Mo, Man Yue}, booktitle = {Proceedings of the NIPS 2014 Workshop on High-energy Physics and Machine Learning}, pages = {129--134}, year = {2015}, editor = {Cowan, Glen and Germain, Cécile and Guyon, Isabelle and Kégl, Balázs and Rousseau, David}, volume = {42}, series = {Proceedings of Machine Learning Research}, address = {Montreal, Canada}, month = {13 Dec}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v42/mack14.pdf}, url = {https://proceedings.mlr.press/v42/mack14.html}, abstract = {We introduce a minorization-maximization approach to optimizing common measures of discovery significance in high energy physics. The approach alternates between solving a weighted binary classification problem and updating class weights in a simple, closed-form manner. Moreover, an argument based on convex duality shows that an improvement in weighted classification error on any round yields a commensurate improvement in discovery significance. We complement our derivation with experimental results from the 2014 Higgs boson machine learning challenge.} }
Endnote
%0 Conference Paper %T Weighted Classification Cascades for Optimizing Discovery Significance in the HiggsML Challenge %A Lester Mackey %A Jordan Bryan %A Man Yue Mo %B Proceedings of the NIPS 2014 Workshop on High-energy Physics and Machine Learning %C Proceedings of Machine Learning Research %D 2015 %E Glen Cowan %E Cécile Germain %E Isabelle Guyon %E Balázs Kégl %E David Rousseau %F pmlr-v42-mack14 %I PMLR %P 129--134 %U https://proceedings.mlr.press/v42/mack14.html %V 42 %X We introduce a minorization-maximization approach to optimizing common measures of discovery significance in high energy physics. The approach alternates between solving a weighted binary classification problem and updating class weights in a simple, closed-form manner. Moreover, an argument based on convex duality shows that an improvement in weighted classification error on any round yields a commensurate improvement in discovery significance. We complement our derivation with experimental results from the 2014 Higgs boson machine learning challenge.
RIS
TY - CPAPER TI - Weighted Classification Cascades for Optimizing Discovery Significance in the HiggsML Challenge AU - Lester Mackey AU - Jordan Bryan AU - Man Yue Mo BT - Proceedings of the NIPS 2014 Workshop on High-energy Physics and Machine Learning DA - 2015/08/27 ED - Glen Cowan ED - Cécile Germain ED - Isabelle Guyon ED - Balázs Kégl ED - David Rousseau ID - pmlr-v42-mack14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 42 SP - 129 EP - 134 L1 - http://proceedings.mlr.press/v42/mack14.pdf UR - https://proceedings.mlr.press/v42/mack14.html AB - We introduce a minorization-maximization approach to optimizing common measures of discovery significance in high energy physics. The approach alternates between solving a weighted binary classification problem and updating class weights in a simple, closed-form manner. Moreover, an argument based on convex duality shows that an improvement in weighted classification error on any round yields a commensurate improvement in discovery significance. We complement our derivation with experimental results from the 2014 Higgs boson machine learning challenge. ER -
APA
Mackey, L., Bryan, J. & Mo, M.Y.. (2015). Weighted Classification Cascades for Optimizing Discovery Significance in the HiggsML Challenge. Proceedings of the NIPS 2014 Workshop on High-energy Physics and Machine Learning, in Proceedings of Machine Learning Research 42:129-134 Available from https://proceedings.mlr.press/v42/mack14.html.

Related Material