ce-regression-intervals
Configure and interpret CE regression intervals for percentile conformal and thresholded probabilistic modes.
npx skills add majiayu000/claude-skill-registry --skill ce-regression-intervals --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# CE Regression Intervals You are configuring or interpreting regression intervals in CE. This skill covers the three distinct regression modes defined in ADR-021. A fully fit + calibrated `WrapCalibratedExplainer` (regression model) is required. Load `references/difficulty_estimator_guide.md` for per-instance adaptive intervals using `DifficultyEstimator`. ## Three Modes — Decision Tree ``` Is threshold= provided? ├── NO -> Mode 2: Percentile conformal intervals (CPS) │ Use low_high_percentiles=(low%, high%) │ Result: interval is a percentile band (e.g., 5th-95th percentile) │ └── YES -> Mode 3: Thresholded probabilistic regression (CPS + Venn-Abers) Result: calibrated P(y <= threshold) with probability interval Interval semantics = probability mass (NOT a percentile band) ``` Mode 1 (classification) is handled by `ce-pipeline-builder` and `ce-factual-explain`. --- ## Mode 2: Percentile Conformal Intervals **Use when:** you want a confidence band around the regression prediction. ```python explanations = explainer.explain_factual(X_query) # default 5th-95th, 90% condfidence explanations = explainer.explain_factual(X_query, low_high_percentiles=(10, 90)) # tighter, 80% confidence e
- Three Modes — Decision Tree
- Mode 2: Percentile Conformal Intervals
- Mode 3: Thresholded Probabilistic Regression
- Calibration Engine (Context — ADR-021)
- Choosing Interval Width
- Alternatives in Regression
- Out of Scope
- Evaluation Checklist
What does the ce-regression-intervals skill do?
Configure and interpret CE regression intervals for percentile conformal and thresholded probabilistic modes.
How do I install it?
Run `npx skills add majiayu000/claude-skill-registry --skill ce-regression-intervals --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From majiayu000/claude-skill-registry, a repository with 534 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.