Writing
Notes & demos
Practitioner notes on applied machine learning, statistics, and data science, usually with something interactive to play with.
- Jun 27, 2026 RAG gives you the machinery for grounding. It does not give you grounding. My coding assistant cited the correct source document and still returned the wrong answer... Why retrieval is not grounding, and what to check when a cited answer is wrong. Interactive demo →
- Jun 9, 2026 Choosing a threshold is a product decision, not a math problem A model gives you scores; the threshold is where you decide what to do with them. An interactive look at the precision and recall tradeoff, and why champion-challenger comparisons go wrong at the default threshold. Interactive demo →
Peer-reviewed
- 2026 Adapting Natural Language Processing Models Across Jurisdictions: A Pilot Study in Canadian Cancer Registries arXiv ↗
- 2025 Small or Large? Zero-Shot or Finetuned? Guiding Language Model Choice for Specialized Applications in Healthcare Machine Learning and Knowledge Extraction ↗
- 2024 Classifying Tumor Reportability Status From Unstructured Electronic Pathology Reports Using Language Models JCO Clinical Cancer Informatics ↗
- 2023 Addressing Inequity in Spatial Access to Lung Cancer Screening Current Oncology ↗
- 2022 Small Area Disease Mapping of Cancer Incidence Using Bayesian Spatial Models Frontiers in Oncology ↗
Full list on Google Scholar.