Interested in Artificial Intelligence, Machine Learning, and the intersection of technology and society.

I enjoy understanding how things work, building practical solutions, and exploring how emerging technologies can create meaningful real-world impact.

• 🏫 Third-year undergraduate at The University of Tokyo, Faculty of Engineering, Program for Systems Innovation (PSI)
• 🔬 Exploring AI/ML, data science, information and communication technologies, and energy systems
• 🎯 Planning to pursue graduate studies and deepen my research in AI
• 🌏 Interested in international research collaboration and cross-cultural perspectives
• 💡 Particularly interested in bridging technical innovation and real-world implementation

• Machine Learning & Deep Learning — Computer vision, NLP, LLMs, model evaluation, and optimization
• AI Systems — Understanding model architectures, inference, and efficient computation
• Information & Communication Technologies — Networks, operating systems, and intelligent systems
• AI for Science — Applying machine learning to scientific and real-world problems
• Technology & Society — Human-centered AI and the societal implications of emerging technologies

🩻 Medical Image Analysis | RSNA Knee

Developing deep learning pipelines for knee MRI classification using CNN-based architectures, including RadNet, ConvNeXt, and CoAtNet.

• Achieved an out-of-fold AUC of approximately 0.944
• Investigated generalization gaps, missing imaging slices, and inference robustness

🤖 LLM Attribution | Name That LLM

Exploring whether language models can be identified from their generated text.

• Combining stylistic features, TF-IDF, LinearSVC, and LightGBM
• Experimenting with Qwen-based LoRA fine-tuning and ensemble methods
• Investigating model-specific writing patterns and potential data leakage

🧪 Molecular Structure Identification | CASMI

Working on molecular structure identification from mass spectrometry data.

• Exploring candidate generation and molecular ranking
• Improving candidate recall and evaluating chemical consistency

🚢 Maritime Data Analysis | AIS

Analyzing vessel movements and predicting passage through the Strait of Hormuz.

• Working with vessel attributes and AIS data
• Emphasizing robust validation and leakage-aware feature engineering

Languages

Python · Java · JavaScript · TypeScript · SQL

Machine Learning & Data Science

PyTorch · scikit-learn · LightGBM · Transformers · LoRA · Pandas · NumPy

Web Development

React · Vite · Node.js · Firebase

Tools & Infrastructure

Git · GitHub · Docker · MySQL · Linux · macOS · Jupyter · Google Colab

Technology is only one part of what interests me. I also enjoy working with people, organizing communities, and exploring different cultures.

• Student Leadership — Served as president of a university student dormitory's self-governing association
• Media & Communication — Worked on student-focused media, content creation, and industry collaboration
• International Experience — Participated in programs at Tsinghua University and in Silicon Valley
• Science Education — Led a prefectural team in Japan's national high school science competition, placing 6th in the experimental category
• Tourism & Communication — Worked as an English-speaking tour guide in Tokyo

Technology should serve people, not the other way around.

I'm interested not only in what AI can do, but also in what we should do with it.

I believe the most interesting challenges lie at the intersection of engineering, human behavior, and society.

🌐 Portfolio: toshiie-shiomi.vercel.app

💻 GitHub: github.com/toshtosh2024

📊 Kaggle: kaggle.com/tosh17

Always curious, always learning.