U.S. flag

An official website of the United States government

Skip Header


Accelerating Entropy Balancing Sample Weighting

Written by:
Working Paper Number: SEHSD-WP2025-05

Survey calibration has proved useful in increasing the accuracy of estimates in practical applications, but the associated computational costs can lead to difficulties with large data sets. In this project, I derive and implement new numerical optimization routines to weight samples using an Entropy Balance calibration approach. This set of algorithms and their associated Python package include extensions of the problem to sparse data, collinear observables, and infeasible moments. Computational results show significant speedup over the widely-used “ebal” R package.

Page Last Revised - May 20, 2025
Is this page helpful?
Thumbs Up Image Yes Thumbs Down Image No
NO THANKS
255 characters maximum 255 characters maximum reached
Thank you for your feedback.
Comments or suggestions?

Top

Back to Header