📉  Predicting Percentiles

Predicting Percentiles#

Forecasting specific percentiles to convey the outcome distribution.

Important

✨ AI-generated content. This page was written with the assistance of an AI language model and is provided as a learning aid. Despite careful review, it may still contain mistakes, omissions, or out-of-date information. Whether you are new to the topic, a team lead, or a senior practitioner, treat it as a starting point rather than an authoritative reference: read it critically and independently verify anything you act on (code, commands, figures, and factual claims) against official documentation and primary sources before relying on it.

What it is#

Predicting percentiles means forecasting specific percentiles (quantiles) of the outcome distribution — the value below which a given percentage of outcomes fall — instead of only a single mean. A percentile is a quantile stated as a percent: the 0.9 quantile is the 90th percentile.

Why percentiles#

A handful of percentiles (say the 10th, 50th and 90th) sketch the range of outcomes and their best- and worst-case scenarios, exposing uncertainty and enabling asymmetric decisions — without committing to a parametric distribution.

How it’s done#

Percentiles are produced by quantile regression (and its tree / boosting variants), each trained on the pinball loss for its level; stacking many percentiles approximates the full distribution.


Theme: Risk & Probabilistic Forecasting  ·  All terminology



See also

Source article Adapted (context, re-expressed) in our own words from: Predicting Percentiles (insightful-data-lab.com).

Tags: purpose: reference topic: terminology level: advanced