Relevance Metrics - Prediction Engine

<div class="section-label">05 Relevance Metrics</div> <div class="section-title">Score relevance without predicting.</div> Sometimes you want the building blocks on their own: how similar each past case is to your circumstances, how informative those cases are, and how they combine into relevance, without forming a full forecast. These scores are what Predict, MaxFit, and Grid use internally when they decide which observations enter the weighted average. For a full forecast, prefer [[Grid Prediction]]. ## Signature ```text relevance(X, theta, cov_inv?) → scores, length N similarity(X, theta, cov_inv?) → scores, length N info_x(X, cov_inv?) → scores, length N # no theta info_theta(X, theta, cov_inv?) → score, length 1 relevance_metrics(X, theta, cov_inv?) → optional: relevance, similarity, info_x, info_theta ``` **Calls:** `relevance`, `similarity`, `info_x`, `info_theta`, `relevance_metrics` **Returns:** score vectors (or a bundle of the scores you request) ## Arguments | Name | Shape | Description | | --- | --- | --- | | X | N × K | Training attributes | | theta | K | Circumstances (not used by info_x) | | cov_inv | K × K | Optional inverse covariance; if omitted, the engine estimates it from X | N = number of observations (rows), K = number of independent variables (columns). With relevance_metrics, request only the scores you need. ## Results Each call returns score vectors you can inspect on their own. The same quantities can also appear on a prediction result under [[Results/Weights & Insights|Weights & Insights]]. | You get | Shape | Look here | | --- | --- | --- | | Similarity | N | How close each training row is to theta | | Informativeness of X (info_x) | N | How unusual each row is vs the sample mean | | Informativeness of θ (info_theta) | 1 | How unusual theta is vs the sample mean | | Relevance | N | Combines similarity with informativeness of the row and of theta | When you run Predict, MaxFit, or Grid, censoring chooses **relevance** or **similarity** (or both on MaxFit and Grid) as the score that decides include vs exclude. ## Related - [[Functions/Overview|Functions]] - [[Predict]] · [[Grid Prediction]] - [[Settings/Options#PredictOptions|PredictOptions]] · [[Results/Weights & Insights|Weights & Insights]]