Skip to content
View woodRock's full-sized avatar
🫖
Activate zen mode (CTRL + K, Z)
🫖
Activate zen mode (CTRL + K, Z)

Highlights

  • Pro

Block or report woodRock

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
woodRock/README.md

profile

📂 whoami — Jesse Wood 👋

Research Assistant @ Victoria University of Wellington | PhD Candidate in Artificial Intelligence

Jesse's GitHub Streak

See You Space Cowboy... 🌌


🔬 Session 01: Featured Research

Revolutionizing Marine Biomass Analysis with Deep Learning & REIMS

Based on my doctoral thesis, this project replaces traditional statistical methods (OPLS-DA) with Transformer-based architectures to analyze Rapid Evaporative Ionization Mass Spectrometry (REIMS) data for the seafood industry.

🚀 Research Breakthroughs

  • Gone Phishing (MoE Transformer): Achieved 100% accuracy in Species Identification (Hoki vs. Mackerel).
  • Autobots (Multi-scale Ensemble): 74.13% accuracy in Body Part Classification—crushing the 51.17% statistical baseline.
  • SpectroSim: A self-supervised contrastive learning framework (SimCLR for chemistry) enabling Batch Traceability without physical tags.
  • Explainable AI (XAI): Utilizing LIME and SHAP to map model decisions back to specific m/z chemical peaks.

💽 The Arsenal: Tech Stack

🧠 AI & Data Science

Python PyTorch HuggingFace Scikit-Learn Pandas NumPy

🧪 Optimization & Explainability

Optuna DEAP XAI WandB

💻 Systems & Engineering

C++ Rust Haskell Docker Streamlit


📜 Bounties: Publications & Demos

Venue/Project Focus Link
PhD Thesis AI for Marine Biomass Research Proposal 📝
IMTS 2025 Electronic Nose Tech Springer Paper 📄
AJCAI 2022 Machine Learning View Paper 📄
Ionic Scholar Academic Tooling Live Demo 🚀

🕹 The Arcade: Game Dev (PICO-8 & Itch.io)

Game Platform Link
Bartender Itch.io Play 🍸
Origins PICO-8 Play 🐍
Pico's Gate PICO-8 Play 🏰
Pico Underlords PICO-8 Play ⚔️
NZ Driving Simulator React Play 🚗

🚬 The Data Cowboy: About Me

I’m a Research Assistant at Victoria University of Wellington. My journey from NIWA software engineering to a PhD in AI is driven by a desire to use code as a scientific medium.

  • 🔭 Current Focus: Self-supervised learning, Masked Spectra Modeling (MSM), and Evolutionary Computation.
  • 👾 Fun Fact: I once built a "Skyrim Wellbeing Manager" to gamify tracking mental health.
  • 🎮 Currently playing: Baldur's Gate 3.

📡 Frequency: Connect

Linktree GitHub

Pinned Loading

  1. fishy-business fishy-business Public

    Machine Learning for Rapid Evaporative Ionization Mass Spectrometry for Marine Biomass Analysis --- A Doctoral Thesis by Jesse Wood

    Python 3 2

  2. qsh qsh Public

    A local-first, privacy-focused CLI that brings vision and semantic understanding to the Unix pipe. Powered by Qwen 3.5-0.8B and Rust. 🦀

    Python 4 3

  3. expert-chainsaw expert-chainsaw Public

    This GitHub Action used the Open Weather API to display the weather forecast for a given area. It is updated once every 30 minutes. The weather forecast is displayed within predefined tags (hidden …

    JavaScript 7 3

  4. ionic-scholar ionic-scholar Public

    This individually developed app keeps track of academic references. The app remembers the users progress, keywords, quotes. Also it can generate citations. We design the app to reduce the stress of…

    TypeScript 3

  5. wordle wordle Public

    Wordle solver with a transformer deep learning neural network.

    Python 1 1

  6. moody-bitch moody-bitch Public

    Moody bitch is a high-fidelity mental health odyssey that transforms real-world self-care into an epic, Skyrim-inspired RPG experience. By completing daily "quests," users earn experience points to…

    TypeScript