liveBook · Manning

Weighted quantile sketch

A weighted quantile sketch is an approximate algorithm for estimating quantiles or quantile-based split points while accounting for item weights.

How 2 Manning books explain it

Ensemble Methods for Machine Learning

A weighted quantile sketch is an approximate data-structuring algorithm for estimating quantile-based split points while accounting for example weights.

How it is used
Speeds XGBoost tree training by approximating feature split selection for large-scale Newton boosting.
How deep it goes
Explains weighted split selection, feature-dependent buckets, and scalability, but defers the considerable implementation details.
Examples
Uses feature-dependent buckets, Hessian-weighted examples, and large-dataset XGBoost training as concrete illustrations.
What you can do after
Evaluate approximate split finding as an efficiency tradeoff in XGBoost and distinguish it from ordinary histogram binning.

Read how Ensemble Methods for Machine Learning explains weighted quantile sketch →

Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI

An approximate quantile algorithm that uses weights, such as gradient statistics, to identify candidate split points.

How it is used
Accelerates gradient-boosted tree split finding by approximating candidate thresholds from weighted gradient statistics.
How deep it goes
Explains weighted candidates, histogram acceleration, and bin reuse, without deriving the sketch’s internal quantile construction.
Examples
Compares exact and 256-bin histogram splitting on latitude using matching thresholds and sharply different runtimes.
What you can do after
Evaluate XGBoost split-finding choices by comparing exact, approx, and hist methods and their bin-reuse behavior.

Read how Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI explains weighted quantile sketch →

What to know first

Related concepts

Questions about weighted quantile sketch

What is weighted quantile sketch?

A weighted quantile sketch is an approximate algorithm for estimating quantiles or quantile-based split points while accounting for item weights.

Which Manning books explain weighted quantile sketch?

Ensemble Methods for Machine Learning and Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI.

What should I know before learning weighted quantile sketch?

Manning books that teach weighted quantile sketch usually introduce gradient boosting first.

What is the difference between weighted quantile sketch and histogram-based tree learning?

A weighted quantile sketch is an approximate algorithm for estimating quantiles or quantile-based split points while accounting for item weights. Histogram-based tree learning is a decision-tree training technique that groups feature values into discrete bins and searches for splits between bins.

These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.