Curriculum vitae
Jiho Han
AI Research Engineer, SimPlatform Co., Ltd. · Seoul, South Korea
I work on machine learning for heterogeneous time-series systems — interpretable anomaly detection, spatiotemporal modeling, and multimodal fusion, applied to medical and manufacturing domains.
Education
M.S. Artificial Intelligence
University of Michigan–Dearborn · High Distinction, 2024 Honor Scholar
B.S. Computer Engineering
University of Illinois at Urbana–Champaign
Research
Structured Temporal Causality for Interpretable Multivariate Time Series Anomaly Detection
OracleAD — an interpretable unsupervised anomaly detection framework using causal embeddings, stable latent structures (SLS), and dual scoring. State-of-the-art results across multiple real-world datasets.
ENACT-Heart: Ensemble-Based Assessment Using CNN and Transformer on Heart Sounds
A ViT–CNN Mixture-of-Experts framework for heart-sound classification, outperforming single-model baselines.
A Novel Retinal Image Contrast Enhancement: Fuzzy-Based Method
A hybrid FCE–CLAHE preprocessing method for retinal vessel segmentation, outperforming HE, CLAHE, and FCE individually.
Fuzzy Inference System for Simulating Ophthalmologists in the Diagnosis of Diabetic Retinopathy
A fuzzy-logic diagnostic simulator benchmarked against ophthalmologists, validated on the FGADR and APTOS datasets.
A Spatiotemporal Fusion System for Multimodal Data Integration into Time Series and Method Therefor patent filed
A spatiotemporal multimodal fusion framework converting heterogeneous sources into unified time series for downstream learning.
Selected work
AI Coaching System, Korea National Team closed source
Lead software architect for a Python AI coaching package for badminton and para-badminton: full match analysis, player performance tracking, and shuttlecock trajectory estimation.
Minister's Commendation — Ministry of Science and ICT, Republic of Korea
reliable-python
Codex and Claude Code plugin enforcing NASA/JPL Power of Ten guardrails and code-smell review for Python.
Carcass Evaluation Project (KAPE) closed source
Computer-vision module for prototype grading equipment, inferring carcass curvature and cross-sectional area using projected-light geometry.
PyTorch-Microstep
A structured, bite-sized learning path for PyTorch.
Modern-AI-Glossary
A working glossary of contemporary AI terminology.
Teaching
Instructor for Senior Design I/II and lab instructor for Complexity Theory at the University of Michigan–Dearborn (2023) — automata theory, Turing machines, algorithmic complexity, dynamic programming, and NP-completeness.
He possesses a unique talent for clearly explaining complex concepts, making the material accessible to students of diverse backgrounds.
— Dr. Nada Lachtar, Assistant Professor, Western Michigan University