Ethan Spraggon Email & Phone Number

San Diego, Ca, Us

Tutoring math and programming languages like Python and Java through 1-on-1 online sessions

San Diego, California, United States

Developed and executed an end-to-end data science project using AWS cloud services with Jupyter Notebooks to create and deploy machine learning models to analyze proteomics data.• Collaborated with experienced staff to collect, integrate, and prepare research data for analysis.• Automated data cleaning with Python scripts to ensure data quality. • Implemented a random forest model with the R package Caret to extract new insights. • Improved model performance through hyperparameter tuning, reducing training time.• Applied statistical analysis techniques to extract important protein features from data.

San Diego, California, United States

• Deep Learning research under Dr. Talmo Pereira.• Leveraged deep learning for pose estimation and tracking to quantify animal behavior and movement. • Developed a data loader to batch a series of images, randomly augmenting them to improve model performance and prevent overfitting.

San Diego, California, United States

Analyzed protein-compound interaction data to find possible relationships between compound structure and protein degradability for a cheminformatics project.• Maintained and updated local and online databases with accurate and current information. • Developed an interactive R web page to view visualizations and insights from data.• Clustered compounds with a T-SNE plot, highlighting 7 different groups of compounds.• Visualized compound clusters with ggplot2 in R with a hover view.

San Diego, California, United States

• High Energy Particle Astrophysics research under Dr. Thomas Siegert.• Analyzed data collected by the SPI detector on the INTEGRAL satellite to find various phenomena such as positron production rates from classical novae, photon flux and ejecta mass from novae.• Deployed Bayesian Hierarchical models with pystan to analyze satellite measurement data of anti-matter particles in space. • Visualized our data with multiple graphics, including exposure maps, signal-to-noise ratio over time, and posterior distributions.• Configured batch scripts to run Python code that fit the data to our model, which gave up to a 5 time improvement (bearing different model assumptions) over previous measurements• Published a paper for this project to Astronomy & Astrophysics: "Nucleosynthesis constraints through γ-ray line measurements from classical novae. Hierarchical model for the ejecta of 22Na and 7Be"

San Diego, California, United States

Math Tutor for kids in all levels of math from algebra 1 to calculus