Skip to content
View cyrustt's full-sized avatar

Highlights

  • Pro

Block or report cyrustt

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 is supported. 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
cyrustt/README.md

Hi, I'm Cyrus Tsui

I am interested in quantitative research at the intersection of mathematical modeling, statistical learning, optimization, and reliable computation.

My projects range from financial econometrics and probabilistic race modeling to counterfactual retrieval, credit-risk modeling, topological analysis of high-dimensional data, and applied matching systems. I enjoy turning quantitative questions into reproducible experiments and using computational evidence to test where an idea works, fails, or needs stronger assumptions.

Selected work

ProjectFocusMethods
Home Credit Default RiskRanking repayment difficulty from application and historical credit dataLightGBM, relational feature engineering, 5-fold OOF validation, Optuna, SHAP
Jockey PredictorEstimating Hong Kong race outcomes and comparing model probabilities with market oddsNGBoost, probabilistic regression, Monte Carlo simulation, Selenium, GitHub Actions
S&P 500 Pairs Trading ResearchTesting whether cointegrated equity pairs support mean-reversion after costsEngle-Granger, ADF, FDR control, rolling OLS, chronological backtesting
Approximate Counterfactual PreservationReducing named features while preserving retrieval-based counterfactualsExact search, beam search, ablations, non-monotone optimization
Topology of the Million Song DatasetComparing acoustic and behavioral structure in music dataPersistent homology, UMAP, HDBSCAN, graph analysis
Imperial Marriage PactPreference-based matching from survey dataVector similarity, constraints, ranking, batch data pipelines

Research interests

  • Statistical learning and high-dimensional data
  • Probabilistic modeling and Monte Carlo methods
  • Time-series analysis and financial econometrics
  • Counterfactual reasoning and model interpretability
  • Probability, numerical methods, and algorithmic reasoning

Tools

Python · NumPy · pandas · SciPy · statsmodels · scikit-learn · LightGBM · NGBoost · Optuna · SHAP · scientific computing · experiment design · data visualization

Elsewhere

Pinned Loading

  1. counterfactual-preservationcounterfactual-preservationPublic

    Research toolkit for preserving retrieval-based counterfactuals under feature reduction

    Python

  2. cyrustt.github.iocyrustt.github.ioPublic

    HTML

  3. Imperial-Marriage-PactImperial-Marriage-PactPublic

    Marriage Pact is a stable matching algorithm based on the Gale-Shapley algorithm to match people with their soulmates.

    Python

  4. TDA_projectTDA_projectPublic

    Jupyter Notebook