# Ziyang Zhou (周梓洋)

> Applied Machine Learning · Statistical Modeling | UBC Statistics. BSc expected May 2027.

I'm Ziyang Zhou (周梓洋), a fourth-year Statistics undergraduate at the University of British Columbia (UBC), expecting to graduate in May 2027. My strengths are applied machine learning, statistical modeling and model evaluation. I work across the modeling process: defining prediction tasks, engineering features, training models and testing how well they generalize.

我是周梓洋（Ziyang Zhou），英属哥伦比亚大学（UBC）统计学专业大四本科生，预计于 2027 年 5 月毕业。我擅长应用机器学习、统计建模与模型评估，实践覆盖预测任务定义、特征工程、模型训练及泛化能力验证。

## Profile pages

- [English home](https://ziyangzhou.me/): Machine learning focus and selected work.
- [中文首页](https://ziyangzhou.me/zh/): 周梓洋、UBC 统计学与机器学习。
- [About](https://ziyangzhou.me/about/): Background, strengths and technical toolkit.
- [Experience](https://ziyangzhou.me/experience/): Internship roles and UBC education.
- [Projects](https://ziyangzhou.me/projects/): Complete project collection.
- [CV](https://ziyangzhou.me/resume/): Current resume.
- [Contact](https://ziyangzhou.me/contact/): Opportunities and collaboration.

## Project case studies

- [CSI 300 Portfolio Modeling Challenge](https://ziyangzhou.me/projects/csi300-portfolio-modeling-challenge/): A project for the 2026 China Collegiate Computing Contest Big Data Challenge. It predicts short-horizon returns for CSI 300 candidates and produces a five-stock allocation. The main work was preparing time-aligned features, comparing validation periods, and deciding whether extra models or features improved the result.
- [Neural Debris Removal Competition](https://ziyangzhou.me/projects/neural-debris-removal/): A computer-vision competition project using a supplied RetinaNet detector. I compared unlearning and detection-filtering approaches, checked what each change removed, and reviewed how much normal detection behavior it preserved.
- [ZZYAgent: Local Coding Assistant](https://ziyangzhou.me/projects/deepseek-coding-agent/): A personal prototype for learning how a coding assistant connects model responses to actions in a local project. The updated version shares one runtime across its command line and terminal interface, and adds saved sessions, structured plans, context budgeting, and checks before reporting a task as complete.
- [Tencent AI-HR Growth Demo](https://ziyangzhou.me/projects/tencent-ai-hr-growth-quest/): A Tencent AI-HR demo project focused on 90-day newcomer growth planning. I used it to practice multi-agent collaboration, AI-assisted deployment, basic data persistence, and turning a product idea into a reviewable demo.
- [Airbnb Listing Price Modeling](https://ziyangzhou.me/projects/airbnb-price-modeling/): A notebook-based tabular modeling project for predicting New York City Airbnb listing prices, covering task framing, train/test split, EDA, feature engineering, model training, and evaluation.
- [Local Qwen Chinese Fiction Pipeline](https://ziyangzhou.me/projects/fanqie-qwen-writing-pipeline/): A Python workflow for Chinese fiction generation with Qwen. It separates planning from drafting and passes character notes, prior events, and recent text into the next segment. Saved progress and lightweight text checks make it easier to review or resume a draft.
- [F1 Pit Stop Prediction](https://ziyangzhou.me/projects/f1-pit-stop-prediction/): A Kaggle Playground project for predicting next-lap pit stops. I organized it as a small experiment pipeline with race-aware cross-validation, saved out-of-fold predictions, comparable configurations, and a checked submission-writing step rather than relying on one notebook run.
- [NLP Text Analysis and Topic Modeling](https://ziyangzhou.me/projects/text-embeddings-topic-modeling/): A self-written CPSC 330 project combining GloVe embeddings, word similarity analysis, spaCy preprocessing, and LDA topic modeling. I explored how text representations affect the patterns and topics a model finds.
- [Weekly Price Forecasting](https://ziyangzhou.me/projects/avocado-price-forecasting/): A time-series coursework exercise using weekly avocado prices across regions and product types. I constructed next-week targets and lag features, used a chronological train/test split, and compared Random Forest predictions with a baseline that reuses the current price.
- [A-Share Data Analysis Tools](https://ziyangzhou.me/projects/astock-analysis-toolkit/): A small collection of AI-assisted A-share data tools. The two Streamlit apps organize intraday transactions, technical indicators, basic fundamentals, flow estimates, and risk views; the companion checker parses AKShare's stock documentation and tests endpoints with timeouts, retries, resume support, and structured reports.
- [ZiyangZhou.me Portfolio Website](https://ziyangzhou.me/projects/ziyangzhou-me-website/): A personal portfolio site that organizes my projects, GitHub repositories, project write-ups, resume content, and bilingual presentation into one recruiter-facing entry point.
- [ChatGPT API Browser Bridge](https://ziyangzhou.me/projects/chatgpt-local-api-creator/): A local API/browser bridge experiment for connecting local scripts with ChatGPT web workflows, especially for long-form generation tasks where repeated context entry becomes inefficient.
- [Recipe Text Clustering](https://ziyangzhou.me/projects/recipe-clustering/): An unsupervised text clustering project comparing bag-of-words and sentence embeddings for recipe-name clustering, using K-means, DBSCAN, and UMAP visualization.
- [Adult Census Preprocessing Pipeline](https://ziyangzhou.me/projects/adult-census-preprocessing-pipeline/): A supervised learning pipeline project on the Adult Census dataset, focused on feature type separation, imputation, encoding, scaling, leakage prevention, and baseline comparison.
- [CPSC 221 Hash Tables Lab](https://ziyangzhou.me/projects/cpsc221-hash-tables/): A UBC CPSC 221 C++ lab focused on hash-table implementation. The folder includes separate chaining, linear probing, iterators, resizing behavior, and text-analysis applications such as word counting and anagram finding.
- [CPSC 221 AVL Trees Lab](https://ziyangzhou.me/projects/cpsc221-avl-trees/): A UBC CPSC 221 C++ lab for AVL tree implementation. The folder contains avltree.cpp, avltree.h, tests, expected outputs, and tree-printing utilities for checking rotation and balancing behavior.
- [CPSC 221 Heaps Lab](https://ziyangzhou.me/projects/cpsc221-heaps/): A UBC CPSC 221 C++ lab focused on array-backed heap implementation and heap operation tests. The work centers on heapify, upheap/downheap behavior, priority-queue ordering, and index arithmetic.
- [CPSC 221 Graphs and Union-Find Lab](https://ziyangzhou.me/projects/cpsc221-graphs-union-find/): A UBC CPSC 221 C++ graph lab with graph.cpp, graph_tools.cpp, edge structures, premade graph utilities, tests, and disjoint-set files. It focuses on representation, connectivity, weighted edges, and union-find reasoning.
- [Sleep Tracking App](https://ziyangzhou.me/projects/sleep-app/): A small Java app built around the personal problem of recording sleep consistently, comparing it with goals, and reviewing patterns over time.
- [China Glass Futures Quantitative Strategy](https://ziyangzhou.me/projects/glass-futures-trading-strategy/): A quantitative research project for China glass futures that reviews strategy logic, backtest behavior, market regimes, drawdown, trade frequency, and parameter stability instead of only highlighting returns.
- [Logistic Regression Text Classification](https://ziyangzhou.me/projects/logistic-regression-text-classification/): A tweet text classification project that moves from a dummy baseline to logistic regression with bag-of-words features, hyperparameter tuning, and coefficient interpretation.
- [Spotify Decision Tree Modeling](https://ziyangzhou.me/projects/spotify-decision-tree-modeling/): A compact decision-tree classification exercise focused on model training, cross-validation, and readable tree visualization.
- [Handwritten Character Recognition App](https://ziyangzhou.me/projects/english-letter-recognition/): A small handwritten-character recognition application built with EMNIST and PyTorch. It separates dataset loading, CNN training, evaluation, prediction, image preprocessing, and the Pygame drawing interface, while keeping compatibility with an earlier saved model.

## Public profiles

- [GitHub](https://github.com/Asdzzyandzzy)
- [GitHub Pages](https://asdzzyandzzy.github.io/)
- [LinkedIn](https://www.linkedin.com/in/ziyang-zhou-a9a16a391/)

## Crawl access

Public pages welcome search engines, AI crawlers and other web crawlers. No sign-in is required. The same content is served to visitors and crawlers.
- [robots.txt](https://ziyangzhou.me/robots.txt)
- [XML sitemap](https://ziyangzhou.me/sitemap.xml)
- [Site index](https://ziyangzhou.me/sitemap/)