🎓 4th year undergraduate student, Department of Industrial Engineering, Hanyang University

📌 Graduating soon and preparing for graduate-level research

🧪 Ongoing research:
“Enhancing commodity factor strategies with deep learning: evidence from basis-momentum” (Submitted for international publication)

https://github.com/SOOBEENKIM/Commodity-Factors.git

💡 Interested in:

• Investment Science, Time Series Forecasting, Data Mining
• Artificial Intelligence (AI), Cybersecurity, Data Analytics
• Financial Engineering, Optimization, Operations Research

💻 ORCID https://orcid.org/0009-0004-7688-1921

🧾 Enhancing Commodity Factor Strategies with Deep Learning
Applied LSTM and Transformer models to dynamically rank 19 commodity futures based on basis-momentum (BMOM), a hybrid factor combining momentum and term structure. https://github.com/SOOBEENKIM/Commodity-Factors.git https://submission.wiley.com/submission/submissionBoard/a04a9c0d-b103-4f8b-825b-c716ef4bc08d/finalReview

🧾 Deep Learning-Based Malware Generation and Classification Developed an LSTM-based sequence autoencoder to extract latent vectors from malicious Python code across 11 behavioral types. Trained a conditional GAN (CGAN) to generate malware-type-specific latent vectors, which were decoded back to source code using autoencoder decoders. Evaluated classification accuracy, reconstruction error, and semantic consistency of generated samples. https://github.com/SOOBEENKIM/Malware_Classification-and-Generation.git

🧾 Impact of Audit Reform on MD&A Tone and Performance Deviations
Conducted a text mining analysis to quantify tone shifts in MD&A before and after Korea’s New External Audit Act. Used Difference-in-Differences methodology to show significant increases in abnormal tone post-reform, suggesting improved strategic language use and enhanced investor transparency.

🧾 Volatility Prediction for Bitcoin using LSTM-GARCH Hybrid Model
Integrated LSTM and GARCH models to jointly predict return and volatility of Bitcoin. Demonstrated improved forecasting accuracy under high volatility regimes. https://github.com/SOOBEENKIM/LSTM-GARCH.git

🧾 Experimental Design Using MINITAB: Cold-Temperature Adhesion Study
Applied a 2⁷⁻³ fractional factorial DOE to analyze effects of phosphoric acid deoxidizer, neutralizer concentration/time, and adhesion promoters on aluminum bonding strength at –65°F. Identified significant factor interactions impacting low-temp peel strength.

🌱 Impact Business Project: Coffee Grounds Upcycling
Developed a solution to encourage coffee shops to participate in spent grounds collection through a subscription model and eco-friendly platform. The initiative aimed to raise public awareness of coffee waste as a recyclable resource.

🧪 Steam R&E : Microbiome Analysis in Fermented Foods
Investigated microbial diversity changes in Korean fermented foods (cheonggukjang, jeotgal, kimchi) across fermentation periods using 16S rRNA sequencing via Ion Torrent and Ion Reporter pipelines. Results showed pH/salinity-driven shifts in microbiome structures. https://github.com/SOOBEENKIM/Microbiome-Analysis.git

🧪 Frontier Chemistry: Lithium Extraction from Seawater
Designed ion-exchange resins using plant-based cellulose to selectively absorb lithium ions from seawater. Compared Li absorption efficiency across different functional groups and bead formation conditions.

🧪 STEAM R&E : Gill-Inspired Fine Dust Capture Device
Designed a cross-step filtration system modeled after fish gill structures to capture fine dust in vehicle exhaust. Used SolidWorks and Flow Design to simulate fluid flow, and built a detachable .STL-based prototype optimized for safety and energy efficiency. https://github.com/SOOBEENKIM/DustCapture-Device.git

🧪 Acoustic Phase Control for Directional Speakers (Physics)
Developed a directional sound system using acoustic phase control. Applied MATLAB to optimize speaker array configurations and control signal origin with function generator input.

💻 Q-Learning in Maze Navigation (Numerical Analysis Project)
Explored the effect of discount factor (gamma) on Q-learning convergence when solving a maze. Analyzed exploration-exploitation tradeoffs using MATLAB; found minimal difference in convergence time on small maps. https://github.com/SOOBEENKIM/Q-Learning.git

📊 Project Management Engineering – Seminar Presentation
Delivered a presentation on “Make Megaprojects More Modular”, discussing modularization benefits in large-scale engineering projects.

💻 OOP Projects – Game Development in Java
Built generalized rock-paper-scissors game and a “catch-the-mouse” strategy game in Java as part of team assignments in object-oriented programming.

🤖 Digital Literacy Startup Proposal using ChatGPT
Proposed a digital literacy education service powered by ChatGPT. Designed curriculum frameworks targeting students and non-technical users. https://github.com/SOOBEENKIM/DigitalLiteracy.git

💻 Time Series Forecasting: ARMA for Air Passenger Demand
Forecasted monthly air passenger volume using ARMA modeling. Evaluated forecasting accuracy and seasonality components.

💻 Financial Engineering Projects
Modern Portfolio Theory, CAPM and Factor Model, Closed-form BlackScholesMerton pricing model, Binomial tree https://github.com/SOOBEENKIM/FinancialEngineering1.git / https://github.com/SOOBEENKIM/FinancialEngineering2.git

💻 P2P Loan Approval Prediction (SVM Model)
Built a classification model using SVM to predict loan approval decisions based on borrower-level financial attributes from real-world P2P lending data. https://github.com/SOOBEENKIM/P2P_Loan.git

💻 Used Car Price Prediction (Ensemble Models)
Applied Random Forest and AdaBoost regression to predict used car market prices. Preprocessed real-world listings and evaluated performance using RMSE. https://github.com/SOOBEENKIM/UsedCar_Adaboost.git