Ruthwik Dhama
From exoplanet enthusiast to creator of a novel machine learning habitability framework
Where Ruthwik Started
His Background
- • 11th grader with strong physics foundation (AP Physics C, Calc III, Differential Equations)
- • Previous research with UNC-G professor on stellar periodicity
- • Grand award winner and gold medalist at international astronomy competition
- • Experience with IRAF and Gaussian curve modeling
His Goals
- • Publish paper in prestigious journals
- • Win Regeneron STS
- • Contribute meaningfully to the field of astrophysics
- • Build a foundation for future research
Prior Research Experience
Before YRI, Ruthwik had already conducted research with a professor from UNC-Greensboro, discovering a novel periodicity for a supergiant star. This work was presented at the NC Astronomers' Meeting and won grand award and gold medal at an international competition.
His Initial Research Idea
"I want to use a model that analyzes how habitable K2-18b is with its recent discovery by JWST of DMS and DMDS compounds, as they could hold life."
— Ruthwik, before starting the program
The Research
Working with his YRI mentor, Ruthwik expanded his initial idea into something far more ambitious: creating an entirely new habitability index for exoplanets. Instead of just analyzing one planet, he built METHI (Machine Learned Exoplanetary Habitability Index)—a novel, data-driven framework that improves upon existing methods like ESI, PHI, and SEPHI.
What Makes METHI Novel
Existing habitability indices (ESI, PHI, SEPHI) rely on fixed heuristics and miss non-linear interactions
Data-driven ensemble learning that captures complex relationships in multi-dimensional data
Binary classification, unsupervised clustering, and ensemble-based regression
Achieved 0.903 score and identified top 10 habitable exoplanet candidates
Public Web Interface
Beyond the research, Ruthwik built a publicly accessible web interface that allows users to input planetary names and retrieve real-time habitability scores—making his research directly usable by the scientific community.
The Outcome
METHI: An Ensemble-based Machine Learned Exoplanetary Habitability Index
2025 8th International Conference on New Media Studies (CONMEDIA)
October 14-17, 2025
First Author
Wanted to analyze habitability of one exoplanet (K2-18b)
Created a novel ML framework that scores all exoplanets, published in IEEE
Ready to Start Your Research Journey?
Join the YRI Fellowship and work with expert mentors to publish your own research.
Apply Now