Columbus, OH
Michael Blevins
Aspiring astrophysicist, software developer, and data scientist applying computational tools and emerging technologies to develop novel, innovative solutions to complex problems.
About
EDUCATION
I graduated from The Ohio State University with a degree in Astronomy & Astrophysics. During my time as a student at The Ohio State University, I developed excellent capability in astronomical data analysis as well as a deep understanding of the underlying statistics, providing me with the framework to succeed in my future endeavors as I work towards becoming an astronomer.
HOBBIES & INTERESTS
MUSIC
Music is deeply special to me; it is so beautifully human. There is no other thing in life that can evoke such raw, vigorous emotions and feelings of passion, sorrow, heartache, joy, ecstasy—the way music does. No matter what you're feeling or where you're at in life, there is a song—a heavenly collaboration of musical instruments and human voices that can ease your suffering, or amplify your happiness, and fill you with life.
I like all music, and I think there is something to appreciate about every song. Below you can find some of my favorite songs and albums.
Favorite Album
This is one of my favorite albums of all time, from my favorite band of all time:
Rainbow Kitten Surprise.
Give it a listen.
Some of my favorite songs from it are
Cold Love,
Bare Bones,
Goodnight Chicago, and
All's Well That Ends.
Favorite Songs
I love sharing music and talking about music with people, and finding new music! Please feel free to send me some of your favorite songs or recommend me an artist I should listen to, I would genuinely love to listen to them.
My Goal
My goal is to contribute to pushing the frontier of our understanding of the universe, whether through astronomical research, scientific communication and outreach, or supporting scientists in various ways, such as processing and archiving data, aiding in software pipeline development, or running observations with astronomical equipment.
- Working on projects to apply, enhance, and expand my skillset to increase my versatility as a scientist.
- Continuing my learning independently through lectures on modern astronomical research practices and working through interactive textbooks in Jupyter.
- Getting involved. Attending science talks, participating in clubs/societies, reaching out and making connections with people in the field.
Projects
Academic Research
Unsupervised Discovery of Galactic Substructures and Anomalies in the JWST COSMOS-Web Survey
I combed through James Webb Space Telescope galaxy catalogs to find anomalous sources that potentially merit further scientific follow-up.
I developed an automated clustering pipeline in Python to analyze data for over 90,000 galaxies. By interpreting cluster physical fingerprints through Z-score heatmaps, the clustering model revealed novel data anomalies, including rare dwarf galaxies and metal-poor starbursts.
Machine Learning for Dwarf Satellite Detection
We developed a binary classifier to identify rare dwarf satellite galaxies in the SAGA Survey DR3, utilizing cross-modal embeddings and a training set of photometrically indistinguishable background sources to capture morphological features beyond standard photometric data.
We further applied dimensionality reduction techniques and trained a logistic regression model, optimizing the decision threshold to achieve 95% recall to maximize the recovery of potential candidates for scientific follow-up.
AI Engineering
Blevins AI Tutor
I'm currently working on a project to build my own personal AI tutor!
I utilize Retrieval Augmented Generation (RAG) to access all of my undergraduate course materials including lectures, textbooks, tutorials, and assignments. I'm building content-specific RAG Agents with LlamaIndex and storing text embeddings in a ChromaDB vector database.
Check it out!Data Analysis & Machine Learning
- Bayesian Statistics
- Linear/Logistic Regression
- Dimensionality Reduction
- Neural Networks (MLP, Convolutional, Bayesian, Siamese)
- Clustering Algorithms
- Retrieval Augmented Generation (RAG)
- Agentic LLMs
Web Development
- React
- Next.js
- Tailwind CSS
- TypeScript
Links
Check out my GitHub and connect with me on LinkedIn!
