Data Scientist & AI Researcher

I’m passionate about tackling diverse challenges and exploring innovative solutions through artificial intelligence. My approach is driven by persistence—I carefully listen to and embrace my colleagues' perspectives, then validate their insights through hands-on analysis.

Constantly striving to evolve as a Data Scientist and AI Researcher, I dive headfirst into every new opportunity. Welcome to my GitHub—feel free to reach out anytime, whether it's morning, afternoon, or even midnight.

Thanks for stopping by!

• LG Aimers 4th / 2024.01.02 ~ 2024.02.26
• LG Aimers 5th / 2024.07.02 ~ 2024.08.30
• LG Aimers 6th / 2025.01.02 ~ 2025.02.27

• LG Aimers 4th Online Hackathon / Public 40th, Private 102th among 800teams
• LG Aimers 5th Online Hackathon / Public 4th, Private 3rd among 740teams
• LG Aimers 5th Offline Hackathon / Public 3rd, Private 2nd among 24teams [finished 3rd overall due to presentation score]
• LG Aimers 6th Online Hackathon / Public 1th, Private 1th among 794teams
• LG Aimers 6th Offline Hackathon / Public 8th, Private 13th among 27teams

• [Classification] Predicting Pregnancy Success in Infertility Patients [with LG Aimers 6th]
• [Classification] Predicting B2B customers who can be converted into sales customers [with LG Aimers 5th]
• [Classification / Abnormality Detection] Display Product Abnormality Detection [with LG Aimers 4th]

• [Abnormaly Detection / VLM] Based on CCTV footage, an innovative video analysis system was proposed/developed which performs anomaly detection and generates logs using a VLM (Visual Language Model) for the detected segments, enabling intelligent situation awareness and automatic reporting.

• [Agent / RAG / Prompt Engineering] For machine learning model interpretation, i developed an LLM architecture that explains SHAP values, based on the activities of the 6th LG Aimers.

• Seo, J.H., Kim, M.S., & Choi, S.H. (2025).
  Factor Analysis with a Framework for Predicting Musical Residual Seats Using Machine Learning.
  Journal of Digital Contents Society, 26(3), 739–747.