EHR Risk 1. First hour
- Home (incl. U.S. healthcare research context)
- Why it matters
- How it works A–Z
- Docker for beginners
- Quickstart
- Workbench hub
Documentation map
Every public page in this site, grouped by learning path. Start anywhere — each page links onward.
Learning paths
All pages
| Page | What you get | Highlights |
|---|---|---|
| Home | Overview + U.S. research context | Leakage · calibrate · SHAP · Docker |
| Why it matters | Problems, audiences, U.S. research relevance | Students · labs · methods |
| Features | Capability catalog | Ingest → trust → serve |
| How it works A–Z | Full walkthrough with screenshots | Home → Predict narrative |
| Workbench hub | All UI screenshots + detail pages | 9 routes covered |
| UI · Home … OpenAPI | Per-page deep dives | See hub table |
| Quickstart | First successful loop | Docker compose · Hub pulls |
| Docker for beginners | Install and verify Docker | Desktop · Engine · Compose |
| Docker Hub images | Pull api + web images | ranasl62/ehr-risk-* |
| Commands | Copy-paste recipes | Make · curl · tests |
| Diagrams | SVG + ASCII architecture | Data flow · temporal |
| Fine-tuning | Model iteration playbook | Compare · promote |
| Architecture | Component map | Jobs · config |
| Data | Schema & integrity | Index / horizon |
| API | Endpoint overview | Predict · jobs |
| Limits | Honest non-goals | Model card · safety |
| Help library | Full guide list + examples | Mirrors in-app Docs |
| Blog | Methods, policy, and research context | 13 posts |
| Compare | vs ad-hoc notebooks | Methods teaching |
| Alternatives | vs opaque AutoML | Research scope |
| Listicle | Open clinical risk tools | Discovery |
| Cite & feedback | Citation + email | support@larucare.com |
Blog series
Methods context and public-source summaries; all posts are research and education only.
Methods & validation
Policy & research infrastructure
Repository companions
Long-form files on GitHub that this site summarizes — open them for thesis appendices.
Core capabilities
Temporal integrity + leakage audit
Features truncated at index time; audit job flags common future-leak mistakes.
Calibration (Brier / ECE)
Optional isotonic calibration on hold-out evaluation.
SHAP in API + UI
Explain drivers on Predict; generate SHAP from Results.
Docker researcher loop
Datasets → Train → Analytics → Predict.