Randomness criteria in terms of α-divergences
نویسنده
چکیده
Vovk’s randomness criterion characterizes sequences that are random relative to two distinct computable probability measures. The uniqueness of the criterion lies in the fact that, unlike the standard criterion based on the likelihood ratio test, it is expressed in terms of a geometrical quantity, the Hellinger distance, on the space of probability measures. In this paper, we generalize the randomness criterion to a wider class of geometrical quantities, the α-divergences with −1 < α < 1. The non-extendibility of the criterion across the boundaries α = ±1 is investigated in connection with the likelihood ratio test and information geometry.
منابع مشابه
Measurement compression with quantum side information using shared randomness
We consider the problem of measurement compression with side information in the one-shot setting with shared randomness. In this problem, Alice shares a pure state with Referee and Bob and she performs a measurement on her registers. She wishes to communicate the outcome of this measurement to Bob using shared randomness only, in such a way that the outcome that Bob receives is correctly correl...
متن کاملOn Clustering Histograms with k-Means by Using Mixed α-Divergences
Clustering sets of histograms has become popular thanks to the success of the generic method of bag-of-X used in text categorization and in visual categorization applications. In this paper, we investigate the use of a parametric family of distortion measures, called the α-divergences, for clustering histograms. Since it usually makes sense to deal with symmetric divergences in information retr...
متن کاملObserved Diffusion Processes
In this paper we propose the use of φ-divergences as test statistics to verify simple hypotheses about a one-dimensional parametric diffusion process dXt = b(Xt, θ)dt + σ(Xt, θ)dWt, from discrete observations {Xti , i = 0, . . . , n} with ti = i∆n, i = 0, 1, . . . , n, under the asymptotic scheme ∆n → 0, n∆n → ∞ and n∆n → 0. The class of φ-divergences is wide and includes several special member...
متن کاملSome Recent Progress in Algorithmic Randomness
Recently there has been exciting progress in our understanding of algorithmic randomness for reals, its calibration, and its connection with classical measures of complexity such as degrees of unsolvability. In this paper, I will give a biased review of (some of) this progress. In particular, I will concentrate upon randomness for reals. In this paper “real” will mean a member of Cantor space 2...
متن کاملNon-flat Clusteringwhith Alpha-divergences
The scope of the well-known k-means algorithm has been broadly extended with some recent results: first, the kmeans++ initialization method gives some approximation guarantees; second, the Bregman k-means algorithm generalizes the classical algorithm to the large family of Bregman divergences. The Bregman seeding framework combines approximation guarantees with Bregman divergences. We present h...
متن کامل