Piece-wise linear isotonic regression

Timo Kuosmanen, Juan F. Monge, José L. Ruiz, Xun Zhou · 2026

Read Piece-wise linear isotonic regression by Timo Kuosmanen, Juan F. Monge, José L. Ruiz, Xun Zhou (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-14943.txt.

Description

Timo Kuosmanen and three others smooth isotonic regression, a tuning-free way to fit rising curves that otherwise comes out as a step function. Steps give no useful marginal values, such as shadow prices or elasticities. Their bilevel optimization fits a continuous, monotonic, piecewise-linear curve to the step estimates, even where the shape is not convex. Simulations show better accuracy in convex and non-convex cases, and a Finnish municipal study shows the payoff.

Catalog record

CREATORTimo Kuosmanen, Juan F. Monge, José L. Ruiz, Xun Zhou
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE48776
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.14943

Browse the full library · About the Tower of Babel