Skip to content
View dnpjr's full-sized avatar

Block or report dnpjr

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
dnpjr/README.md

Daniel Pajer — Applied Modelling, Data & ML

Oxford PhD (DPhil in Mathematical Physics). I build structured, reproducible modelling and data workflows in Python — spanning ML pipelines, strategy research/backtesting, and interactive simulation demos.

I’m interested in roles across quantitative modelling, data science / analytics, and risk / research engineering.

Featured Projects

  • Imbalanced Credit Default ML Pipeline (LogReg / RF / XGBoost)
    End-to-end pipeline: data download → cross-validated model selection (PR AUC) → threshold tuning → plots + metrics.
    Example results: ROC AUC ~0.77, PR AUC ~0.55; recall improved from ~35% → ~59% via imbalance-aware evaluation + threshold optimisation.
    Repo: ml_pipeline

  • Monday Range — Research + Backtest Framework
    Research module quantifies conditional outcomes for a weekly liquidity pattern; backtest module includes risk-based sizing and layered exits.
    Repo: monday_range

  • Drone Radar Interception Simulation (HTML Canvas)
    Lightweight browser demo: real-time simulation loop, radar sweep detection, 1:1 target assignment, interception dynamics; deployable via GitHub Pages.
    Repo: drones

Skills Snapshot

Python (NumPy, pandas, scikit-learn, XGBoost), reproducible experiment structure, basic time-series tooling, plotting (matplotlib).
Also: JavaScript/HTML (interactive demo), some MATLAB/C++ (older projects).

Contact

This profile is a living portfolio: projects and write-ups are being expanded with robustness checks (costs/slippage, walk-forward validation, sensitivity analysis) and additional modelling examples.

Popular repositories Loading

  1. drones drones Public

    Interactive HTML5 canvas simulation of radar detection and drone interception (real-time loop + HUD).

    HTML

  2. monday_range monday_range Public

    Research + backtest framework for a weekly range-based pattern with risk-based sizing, layered exits, and reproducible outputs.

    Python

  3. ml_pipeline ml_pipeline Public

    Reproducible credit-default classification pipeline with imbalance-aware evaluation (PR AUC), threshold tuning, and model comparison (LogReg/RF/XGBoost).

    Python

  4. dnpjr dnpjr Public