Skip to content
View Quaden2307's full-sized avatar
🐙
Noodling
🐙
Noodling

Block or report Quaden2307

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.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

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

Report abuse
Quaden2307/README.md
Caden Sun
Contribution graph for the last year

Hi! I'm a Mathematics student at the University of Waterloo, class of 2030. I'm into Software Engineering, Machine Learning, and AI, and always looking for opportunities to build something great. My past experience includes interning as an AI Engineer at ProCogia, and in Winter 2027 I'm joining Bits in Glass as a Machine Learning Engineer Intern. In my free time I like to play the piano, hoops + eats, and going on walks.

Contact

Email is the best way to reach me: c262sun@uwaterloo.ca. I'm also on LinkedIn, and there's more on my website, cadensun.com.

Projects

Flight Price Predictor estimates what a flight should cost. A daily pipeline has collected over 400,000 fares since May 2026, now across the 300 routes below, and the XGBoost model averages 14.6% error on flights held out from training, against 17.1% for a linear baseline.
Map of the 300 routes between 81 airports that the flight price pipeline collects fares on every day

Cat Classifier decides whether your cat is a ragdoll. It's a ResNet-18 convolutional network fine-tuned on a 1,722-photo dataset and run in your browser through ONNX, so the photo never leaves your device. Answering "not ragdoll" every time would already be 77% accurate, so it trains with a class-weighted loss and is judged on F1, where it scores 0.83 and catches 52 of the 60 ragdolls in the test set. Try it.
A pixel cat with a 3 by 3 window over the edge of its ear, that window enlarged, and the edge map a convolution produces

Chess Engine searches with negamax, alpha-beta pruning and iterative deepening, plus quiescence search, a Zobrist-keyed transposition table, killer and history move ordering, and a tapered evaluation. It plays through a Flask API on Vercel and a React board with a live evaluation bar and candidate-move arrows. In the position below White is a queen down, and the engine finds Légal's Mate: 6.Bxf7+ Ke7 7.Nd5#. Play against it.
Chess position with White a queen down; an arrow shows the engine's move Bxf7+, which mates in two

Scout is a lidar rover my team built in 32 hours at Hack the North 2026. It checks the gaps along its path against the 860 mm a wheelchair needs, and its lidar read one doorway at 878 mm against an 880 mm reference. I built the operator console's map view, made the simulator drivable, and added run controls, downloads and reports.

Tools

Languages: Python, C++, C, TypeScript, JavaScript, SQL, Java
Machine learning: PyTorch, scikit-learn, XGBoost, pandas, NumPy, LangGraph, ONNX
Web and cloud: Flask, React, Node.js, AWS, PostgreSQL with pgvector, SQLite, Docker, Linux, Vercel

Pinned Loading

  1. flight-price-predictor flight-price-predictor Public

    Jupyter Notebook 1

  2. cat-classifier cat-classifier Public

    Python

  3. chess-engine chess-engine Public

    Python 1

  4. world-cup-score-prediction-model world-cup-score-prediction-model Public

    Python 1

  5. rank_llm rank_llm Public

    Forked from castorini/rank_llm

    RankLLM is a Python toolkit for reproducible information retrieval research using rerankers, with a focus on listwise reranking.

    Python

  6. student-housing-assistant student-housing-assistant Public

    Python