Samuel May

Postdoctoral researcher at Boston University with a Ph.D. in experimental particle physics. Technical and intellectual leader on multiple highly cited publications on the Higgs boson. My work involves:

• Data engineering: implemented high-throughput computing pipelines for processing O(Petabytes) of data
• Applied ML/AI research: designed ML algorithms including boosted decision trees, recurrent neural networks (LSTMs), and unsupervised algorithms for anomaly detection (PCAs & autoencoders)
• Scientific communication: authored scientific manuscripts ranging from internal documents, publications, and public outreach articles
• Software development: led development of a Python-based framework for measurements of the Higgs boson

Statistical Analysis of the Higgs Boson

• Led analyses of petabytes of particle collision data from the Compact Muon Solenoid (CMS) experiment at the CERN Large Hadron Collider to measure properties of the Higgs boson, resulting in three highly-cited publications in top journals and three talks at major international conferences.
• Developed machine learning (ML) algorithms, including boosted decision trees and recurrent neural networks (LSTMs) which improved the expected statistical significance of a Higgs boson measurement by 10%.

∗ Became recognized as an ML expert within CMS and was selected for conference talk on usage of ML in CMS.

• Embraced both technical and intellectual leadership roles:

∗ Chaired meetings and coordinated efforts of 10+ active collaborators.

∗ Trained and mentored Ph.D. students.

∗ Led authorship of scientific manuscripts.

Software Development

• Led development of a Python framework to serve the 50+ physicists in the CMS Higgs to Diphoton group.
• Pioneered transition from loop-based operations in C++ to columnar-based vectorized operations in Python.
• Implemented job submission and monitoring tools for distributed computing (HTCondor) to process terabytes of data.

Machine Learning Engineering: Data Quality Monitoring

• Led development of a toolkit for using ML algorithms (PCAs and autoencoders) to identify anomalous behavior in the CMS detector.
• Implemented full pipeline: data aggregation and cleaning, training of ML algorithms, assessing and visualizing performance, and tutorial for collaboration-wide use of the tool.
• Supervised Ph.D. students in both contributing to the toolkit and applying it to develop ML algorithms for use with the CMS detector.

Earned Ph.D. in physics (experimental particle physics) from UC San Diego in 2020.

Dissertation: Observation of Higgs boson production in association with a top quark-antiquark pair in the diphoton decay channel

My thesis research was performed on the Compact Muon Solenoid experiment at the Large Hadron Collider, using machine learning and statistical methods to perform measurements of the interaction between the Higgs boson and the top quark.

I was awarded the UC San Diego Physics Department's Harold K. Ticho Fellowship, given to one student each year.