Fitting algebraic curves to noisy data

Sanjeev Arora, Subhash Khot

Research output: Contribution to journalArticlepeer-review

Abstract

We introduce the following problem which is motivated by applications in vision and pattern detection: We are given pairs of datapoints (x1, y1), (x2, y2), ..., (xm, ym) ∈ [-1, 1] × [-1, 1], a noise parameter δ > 0, a degree bound d, and a threshold ρ > 0. We desire an algorithm that enlists every degree d polynomial h such that |h(xi) - yi|≤δ for at least ρ fraction of the indices i. If δ = 0, this is just the list decoding problem that has been popular in complexity theory and for which Sudan gave a poly(m, d) time algorithm. However, for δ>0, the problem as stated becomes ill-posed and one needs a careful reformulation (see the Introduction). We prove a few basic results about this (reformulated) problem. We show that the problem has no polynomial-time algorithm (our counterexample works for ρ = 0.5). This is shown by exhibiting an instance of the problem where the number of solutions is as large as exp(d0.5-ε) and every pair of solutions is far from each other in ℓ norm. On the algorithmic side, we give a rigorous analysis of a brute force algorithm that runs in exponential time. Also, in surprising contrast to our lowerbound, we give a polynomial-time algorithm for learning the polynomials assuming the data is generated using a mixture model in which the mixing weights are "nondegenerate.".

Original languageEnglish (US)
Pages (from-to)325-340
Number of pages16
JournalJournal of Computer and System Sciences
Volume67
Issue number2
DOIs
StatePublished - Sep 2003

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science
  • Applied Mathematics
  • Computer Networks and Communications
  • Computational Theory and Mathematics

Fingerprint

Dive into the research topics of 'Fitting algebraic curves to noisy data'. Together they form a unique fingerprint.

Cite this