Publication Date
2026
Conference/Sponsorship/Institution
40th Conference on Neural Information Processing Systems (NeurIPS 2026)
Description
CMinHash promises a new direction for designing MinHash algorithms with reduced variance and storage space via circulant permutations. However, the original variance analysis is difficult to extend to other MinHash estimators. We present a new approach that gives explicit and easily comparable approximate variance expressions, and use it to analyze the Minner estimator. Numerical experiments show that our variance expression matches the empirical mean square error (MSE) of estimates. Our work thereby demonstrates how to apply circulant permutation to other MinHash estimators and compute their approximate variance.
Type
Conference Paper
Department
Mathematics
Link to published version
https://neurips.cc/virtual/2026/loc/atlanta/poster/153085
Recommended Citation
Kang, Keegan; Hood, Avery; and Wong, Benedict H.W., "Understanding Circulant Permutation To Extend The CMinHash Estimator" (2026). Faculty Conference Papers and Presentations. 106.
https://digitalcommons.bucknell.edu/fac_conf/106