Paper by Erik D. Demaine

Reference:

Ilya Baran and Erik D. Demaine, “Optimal Adaptive Algorithms for Finding the Nearest and Farthest Point on a Parametric Black-Box Curve”, International Journal of Computational Geometry and Applications, volume 15, number 4, 2005, pages 327–350. Special issue of selected papers from the 20th Annual ACM Symposium on Computational Geometry, 2004.

BibTeX

@Article{Curves_IJCGA,
  AUTHOR        = {Ilya Baran and Erik D. Demaine},
  TITLE         = {Optimal Adaptive Algorithms for Finding the Nearest and
                   Farthest Point on a Parametric Black-Box Curve},
  JOURNAL       = {International Journal of Computational Geometry and
                   Applications},
  journalurl    = {http://www.worldscinet.com/ijcga/ijcga.shtml},
  VOLUME        = 15,
  NUMBER        = 4,
  YEAR          = 2005,
  PAGES         = {327--350},
  NOTE          = {Special issue of selected papers from the 20th Annual ACM
                   Symposium on Computational Geometry, 2004.},

  length        = {24 pages},
  papers        = {Curves_SoCG2004},
  replaces      = {Curves_SoCG2004},
  doi           = {https://dx.doi.org/10.1142/S0218195905001737},
  dblp          = {https://dblp.org/rec/journals/ijcga/BaranD05},
  comments      = {This paper is also available from <A HREF="http://www.worldscinet.com/ijcga/15/1504/S0218195905001737.html">WorldSciNet</A>.
                   This paper is also available as
                   <A HREF="http://arXiv.org/abs/cs.CG/0307005">
                   arXiv:cs.CG/0307005</A> of the
                   <A HREF="http://arXiv.org/archive/cs/intro.html">
                   Computing Research Repository (CoRR)</A>.},
  withstudent   = 1,
}

Abstract:

We consider a general model for representing and manipulating parametric curves, in which a curve is specified by a black box mapping a parameter value between 0 and 1 to a point in Euclidean d-space. In this model, we consider the nearest-point-on-curve and farthest-point-on-curve problems: given a curve C and a point p, find a point on C nearest to p or farthest from p. In the general black-box model, no algorithm can solve these problems. Assuming a known bound on the speed of the curve (a Lipschitz condition), the answer can be estimated up to an additive error of ε using O(1/ε) samples, and this bound is tight in the worst case. However, many instances can be solved with substantially fewer samples, and we give algorithms that adapt to the inherent difficulty of the particular instance, up to a logarithmic factor. More precisely, if OPT(C, p, ε) is the minimum number of samples of C that every correct algorithm must perform to achieve tolerance ε, then our algorithm performs O(OPT(C, p, ε) log (ε^-1 / OPT(C, p, ε))) samples. Furthermore, any algorithm requires Ω(k log (ε^-1/k)) samples for some instance C' with OPT(C', p, ε) = k; except that, for the nearest-point-on-curve problem when the distance between C and p is less than ε, OPT is 1 but the upper and lower bounds on the number of samples are both Θ(1/ε). When bounds on relative error are desired, we give algorithms that perform O(OPT ⋅ log (2 + (1+ε^-1) ⋅ m^-1 / OPT)) samples (where m is the exact minimum or maximum distance from p to C) and prove that Ω(OPT ⋅ log (1/ε)) samples are necessary on some problem instances.

Comments:

This paper is also available from WorldSciNet. This paper is also available as arXiv:cs.CG/0307005 of the Computing Research Repository (CoRR).

Length:

The paper is 24 pages.

Availability:

The paper is available in PostScript (5273k), gzipped PostScript (439k), and PDF (0k).

See information on file formats.

[Google Scholar search]

Related papers:

Curves_SoCG2004 (Optimal Adaptive Algorithms for Finding the Nearest and Farthest Point on a Parametric Black-Box Curve)