Research Intern in the Weber Lab
University of Pennsylvania, Neuroscience Department
Developing machine learning models using Pytorch to predict p-waves in brain state data during REM sleep in mice;
Built Convolutional (CNN) and Long Short-Term Memory (LSTM) neural networks from scratch;
Achieved 98% accuracy with an LSTM model and 95% accuracy (RMSE) with a CNN model using 60+ features;
Processed & cleaned 50+ local field potential (LFP) files with > 26 million samples per file;
Generated augmented data; visualized predicted waveforms, accuracy, & loss using Matplotlib
Research Intern under Prof. Charles Yang
University of Pennsylvania, Linguistics Department
Developed an algorithm using Python to mimic how children exercise pattern recognition using the Abductive Discovery of Productivity (ADP) and the Tolerance Principle (TP); Built a recursive decision tree-based algorithm that dynamically resizes based on user input.
CIS 1200 Teaching Assistant
University of Pennsylvania
Programming Languages and Techniques TA; each OCaml, Java, & program design concepts, including functional programming, GUI, & interfaces;
Lead weekly recitation review for 20+ students and office hours for 350+ students;
Develop weekly recitation materials for 50+ TAs including interactive slides & worksheets
Research Intern under Dr. A Surjalal Sharma
University of Maryland, Department of Astronomy
Developed a Long Short-Term Memory (LSTM) recurrent neural network model to predict Geomagnetic Auroral Electrojet Indices using Python;
Achieved 97% accuracy (Root Mean Squared Error);
Presented at the 2022 American Geophysical Union Fall Meeting to 30+ members
CISESS Intern
University of Maryland, Department of Earth & Space Science
Internship under Dr. Yongzhen Fan of NOAA;
Developed a machine learning-based snowfall detection algorithm using Python for NASA’s Global Precipitation Measurement Mission satellite (GPM); Used inputs from 9 microwave sensors;
Achieved 95% classification accuracy using XGBoost with less than 0.1% false prediction rate;
Increased forecast accuracy in Alaska & the Southern Hemisphere from 0% to 94.6%;
Developed XGBoost, Random Forest, & Linear Regression ML models to predict snowfall from 800+ features
ASPIRE Intern
Johns Hopkins Applied Physics Laboratory
Trained a machine learning neural network to identify litter in videos using Tensorflow Lite and Python; Configured the SSD-MobileNet-V2 object detection model on a Raspberry Pi;
Cleaned & labeled 1,500 images of litter; Achieved 90% recall;
Presented at the AIAA Mid-Atlantic Young Professionals, Students, and Educators (YPSE) Conference to 50+ members
Research Intern under Prof. Bengt Eliasson
Virtual
Research intern under Prof. Bengt Eliasson; Studied the “butterfly effect” in climate predictions and weather simulation; Developed MATLAB code to simulate complex dynamic systems including the Mandelbrot set & Lyapunov exponent.
