Skills and Experience

Recent Work Experience

Chase Lab (CMU),
NSF Graduate Research Fellow

Jul. 2018 - Present

Research towards doctoral degree at CMU probing motor cortical population mechanisms of reward-mediated performance.

Relevant skills: multimodal data acquisition, neural population analysis with linear dimensionality reduction, scientific writing and presenting.

Ghandi Lab (Pitt),
CNBC Undergraduate Research Fellow

Jan. 2017 - Mar. 2019

Research on communication in the superior colliculus and dereferencing methods for field potential Geweke-Granger Causality.

Relevant skills: digital filtering, data analysis with Matlab, time-frequency distributions, post-hoc dereferencing schemes.

Philips Respironics, Software Engineering Co-op Intern

Jan. 2016 - Aug. 2017

Back-end web service development on "Care Orchestrator" application for managing patient data and CPAP device use.

Relevant skills: Java and C# for web services, debugging, version control, automated builds and testing, logging, SQL queries.

Expertise

Multimodal data processing

My PhD thesis work has heavily involved the design and implementation of non-human primate experiments that involve recording many streams of data simultaneously. These include:

Neural population spiking activity
Hand position
Gaze position and pupillometry
Gated pulse oximetry for heart rate
Surface electromyography (EMG)

Each data stream has different quirks and challenges for processing, but combined, provide valuable insights into the cognitive processes underlying behavior.

Data analysis

My research mainly focuses on neural population analyses - Using statistical and machine learning methods, I leverage hundreds of simultaneously recorded neurons in the motor cortex of the brain to identify behaviorally-relevant patterns of neural activity. These methods often rely on dimensionality reduction to identify low-D patterns in high-D neural activity.

More generally, I'm experienced with answering quantitative scientific questions with data analysis and statistical methods.

Engineering

...but to even get any of this data, a fair bit of engineering is involved! Simple electrical engineering skills were useful for building basic circuits for signal acquisition (e.g., hand tracking markers, pulse-ox) and hardware/software filtering the signals they output.

I also wrote the software program to run the task shown in the video above from scratch to maximize my ease of use and customizability.

Skills Laundry Lists

Actively using

Matlab, data cleaning and preprocessing, multimodal biosignal acquisition and analysis (multiunit array recordings, hand kinematics, EMG, pulse oximetry, gaze position, pupillometry), digital filtering, neural population analysis techniques, linear dimensionality reduction (PCA, factor analysis), continuous and categorical decoding methods (Fisher/Linear discriminant analysis, naive bayes, population vector algorithm), brain computer interfaces, bootstrapping, statistical testing, research writing, research presentation, creation of publication-quality research figures (Adobe Illustrator), data visualization, critical evaluation of scientific literature, design of primate behavioral experiments, animal training, teaching assistance, tutoring

Previously used

Time-frequency distribution analyses, post-hoc signal dereferencing schemes, autoregressive models, Granger Causality, local field potential analysis, IoT device programming, Java and C# for web services, navigating SQL databases, desktop CNC, 3D-printing, app design with Objective-C

Some experience with

Python for data analysis, Labview, C, nonlinear dimensionality reduction methods, Kalman filters, RNNs, dPCA