• DocumentCode
    109956
  • Title

    Minimum KL-Divergence on Complements of L_{1} Balls

  • Author

    Berend, Daniel ; Harremoes, Peter ; Kontorovich, Aryeh

  • Author_Institution
    Dept. of Math. & Comput. Sci., Ben-Gurion Univ., Beer-Sheva, Israel
  • Volume
    60
  • Issue
    6
  • fYear
    2014
  • fDate
    Jun-14
  • Firstpage
    3172
  • Lastpage
    3177
  • Abstract
    Pinsker´s widely used inequality upper-bounds the total variation distance ∥P - Q∥1 in terms of the Kullback-Leibler divergence D(P∥Q). Although, in general, a bound in the reverse direction is impossible, in many applications the quantity of interest is actually D*(v, Q)-defined, for an arbitrary fixed Q, as the infimum of D(P∥Q) over all distributions P that are at least v-far away from Q in total variation. We show that D*(v, Q) ≤ Cv2 + O(v3), where C = C(Q) = 1/2 for balanced distributions, thereby providing a kind of reverse Pinsker inequality. Some of the structural results obtained in the course of the proof may be of independent interest. An application to large deviations is given.
  • Keywords
    information theory; probability; KL-divergence; Kullback-Leibler divergence; reverse Pinsker inequality; total variation distance; Abstracts; Atomic measurements; Equations; Extraterrestrial measurements; History; Information theory; Standards; McDiarmid´s inequality; Pinsker´s inequality; Sanov´s theorem;
  • fLanguage
    English
  • Journal_Title
    Information Theory, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9448
  • Type

    jour

  • DOI
    10.1109/TIT.2014.2301446
  • Filename
    6746175