Sciweavers

ALT
2015
Springer

Interactive Clustering of Linear Classes and Cryptographic Lower Bounds

7 years 11 months ago
Interactive Clustering of Linear Classes and Cryptographic Lower Bounds
Abstract. We study an interactive model of supervised clustering introduced by Balcan and Blum [6], where the clustering algorithm has query access to a teacher. We give an efficient algorithm clustering linear functionals over finite fields, which implies the learnability of parity functions in this model. We also present an efficient clustering algorithm for hyperplanes which are a natural generalization of the problem of clustering linear functionals over Rd . We also give cryptographic hardness results for interactive clustering. In particular, we show that, under plausible cryptographic assumptions, the interactive clustering problem is intractable for the concept classes of polynomial-size constant-depth threshold circuits, Boolean formulas, and finite automata.
Ádám Dániel Lelkes, Lev Reyzi
Added 15 Apr 2016
Updated 15 Apr 2016
Type Journal
Year 2015
Where ALT
Authors Ádám Dániel Lelkes, Lev Reyzin
Comments (0)