cv

Frontend Developer

Peulda Co., Ltd.

Developed a web application.

• Supervisor: CTO, Hahn Lee

Research Intern

Design & Electronic Convergence System (DECS) Lab in UNIST

Developed 'YoungCHA', an indoor shared mobility platform. Addressed ergonomic limitations of conventional systems by engineering a custom embedded control system for intuitive single-handed driving.

• Supervisor: Prof. Hui-sung Lee
• Embedded System (MBED): Developed firmware using C/C++ to control DC motors and interface with sensors (IMU, Encoders) for the mobility prototype.
• Computer Vision: Integrated vision-based environmental sensing to enhance indoor navigation safety.

Exchange Student

Korea Advanced Institue of Science and Technology (KAIST)

School of Computing

• Introduction to Database
• Introduction to Reinforcement Learning
• Introduction to Artificial Intelligence
• Introduction to Software Engineering
• Automated Software Testing

Bachelor of Science

Ulsan National Institue of Science and Technology (UNIST)

Department of Computer Science and Engineering

• Introduction to Algorithms (A+)
• Computer Networks (A+)
• Operating System (A-)
• Principle of Programming Language (A+)
• Introduction to Data Mining (A+)
• Machine Learning (A)
• Information Visualization (A+)
• Computer Vision (A+)
• GPA: 4.0+/4.30

Grand Prize, 2024 Defense Public Data Utilization Contest

Minister of National Defense, Republic of Korea

Awarded for developing 'Military Welfare Map', a web service integrating scattered military discount information into a map-based UI. Overcame public data limitations by building a custom crawler for 40+ municipal websites and implemented an AI chatbot ('Sergeant GPT') to enhance accessibility for soldiers and their families.

Bronze Prize, 2nd X-Corps Plus Festival

X-Corps Plus

Awarded for the 'YoungCHA' project. Recognized for solving real-world engineering problems regarding steering mechanics and user experience in indoor mobility.

Gold Prize, 2nd U-Challenge Festival

UNIST

Awarded for the 'YoungCHA' project. Recognized for excellence in the practical implementation of an indoor shared mobility system using embedded control technologies.

Best Team (1st) & Best Participant (Top 5), OUTTA AI Bootcamp

OUTTA

Achieved 1st place in the team project competition out of 27 teams. Selected as a Top 5 Best Participant (5th/61) based on individual coding performance and project contributions.

SKiP: SVM weighted by K-Nearest-Neighbors and class Probability for weakening outliers

Jeonghoon Park, Kangjun Lee, Jaemin Kim, Doyeol Oh

Proposed a robust SVM variant using dual confidence metrics to handle both feature and label outliers simultaneously. Combined global class probability (p_i) to detect feature outliers with local KNN consistency (n_i) to identify label outliers, using additive weighting for mutual compensation. Demonstrated consistent improvements over Naive-SVM, Prob-SVM, and KNN-SVM across diverse datasets (Iris, Wine, Titanic) with varying noise levels.

위치 기반 소셜 네트워크에서 공간 근접 커뮤니티 탐지를 위한 엔트로피 가중 적응형 라벨 전파

Doyeol Oh, Hyewon Kim, Dahee Kim, Junghoon Kim

Proposed an adaptive label propagation algorithm for LBSN graph clustering that dynamically balances structural and geographical information using label entropy as a confidence measure. Successfully distinguished geographically distant but structurally connected communities (e.g., Nashville vs. Atlanta) while achieving faster convergence than traditional methods. Accepted at KDBC 2025. Patent application in progress.

Hylos: Hierarchically Localized Optimization Strategy

Doyeol Oh

Proposed a hierarchical divide-and-conquer algorithm combining k-means clustering with dynamic solver selection (Held-Karp for small clusters, Christofides for large ones) to solve large-scale TSP instances. Achieved 7× speedup over traditional Christofides on mona-lisa100k dataset while improving solution quality by 1.71%. Selected as Best Report for CSE331 Introduction to Algorithms Assignment 2, Spring 2025.

실내 공유형 모빌리티의 핸즈프리 구동을 위한 직관적 조향기술 개발

Donghoon Nam, Doyeol Oh, Seongjae Lee, Yoonjoung Kwak, Hui-sung Lee

Developed 'YoungCHA', a chair-type indoor shared mobility platform enabling hands-free driving through saddle rotation instead of traditional handlebar control. Implemented an auto-calibration system using physical limit switches to overcome incremental encoder reset issues, ensuring accurate motor positioning after power cycles. Awarded Gold Prize at 2nd U-Challenge Festival and Bronze Prize at 2nd X-Corps Plus Festival. Accepted at ICROS 2024.