Author(s)
|
Harris, Philip (MIT) ; Mccormack, William Patrick (MIT) ; Park, Sang Eon (MIT) ; Quadfasel, Tobias (Hamburg U.) ; Sommerhalder, Manuel (Hamburg U.) ; Moureaux, Louis Jean (Hamburg U.) ; Kasieczka, Gregor (Hamburg U.) ; Amram, Oz (Fermilab) ; Maksimovic, Petar (Johns Hopkins U.) ; Maier, Benedikt (KIT, Karlsruhe, EKP) ; Pierini, Maurizio (CERN) ; Wozniak, Kinga Anna (CERN) ; Aarrestad, Thea Klaeboe (Zurich, ETH) ; Ngadiuba, Jennifer (Fermilab) ; Zoi, Irene (Fermilab) ; Bright-Thonney, Samuel Kai (Cornell U.) ; David Shih ; Bal, Aritra (KIT, Karlsruhe, EKP) |
Abstract
| We present the performance of Machine Learning--based anomaly detection techniques for extracting potential new physics phenomena in a model-agnostic way with the CMS Experiment at the Large Hadron Collider. We introduce five distinct outlier detection or density estimation techniques, namely CWoLa, Tag N' Train, CATHODE, QUAK, and QR-VAE, tailored for the identification of anomalous jets originating from the decay of unknown heavy particles. We demonstrate the utility of these diverse approaches in enhancing the sensitivity to a wide variety of potential signals and assess their comparative performance in simulation. |