Designed a vertexing algorithm for the future HGC (High-granularity calorimeter) using only spatial and timing information to reconstruct interaction vertex locations to sub-millimeter precision, and applied this to Higgs to di-photon events.

HGCal simulation analyses for CMS - Sarah Marie Bruno

Compared the performance of the current ECAL (Electromagnetic calorimeter) and the future HGC for the Higgs to di-photon channel. Also worked at the CERN test beam with a prototype Shashlik detector.

Used neutral pion events to study ECAL crystal timing response and transparency in 2015 data, including the relationship between the two variables.

Worked at the CERN test beam with a prototype Shashlik detector, and measured the timing resolution and rising edge properties of pulses from a Hamamatsu MCP-PMT (micro-channel plate photomultiplier tube).

Worked at the FNAL test beam with Hamamatsu SiPMs (silicon-based photomultipliers), measured timing resolution with a picosecond-pulsed laser.

Direct tests of a pixelated microchannel plate as the active element of a shower maximum detector - Federico Presutti

Worked at the FNAL test beam to investigate use of Photek/Photonis MCPs in an electromagnetic shower maximum detector.

Trained neural networks for track reconstruction using PyBrain and Theanets libraries.

Trained neural networks to make razor trigger decisions, with focus on fast, hardware- compatible implementations.

Implemented a principle component analysis using data transfer information from the CMS PhEDEx (Physics Experiment Data Export) tool.

Investigated use of SOMs (self-organizing map) for outlier detection in razor searches, and NADEs (Neural autoregressive density estimator) for pseudo-data production.

Investigated use of GPU-based convolution neural networks for a quark-gluon jet discriminator.

Measured jet transverse momentum resolution and bias in pointing reconstruction, hadronic razor trigger efficiency, and electron/muon reconstruction efficiencies in 8 TeV data and 13 TeV Monte Carlo samples for the razor analysis.