Searcharxiv⌕ Search

arXiv subjects

Maxim Raginsky

Publications and source records attributed to Maxim Raginsky.

90 records · Page 5Linked to original sources
arXiv2009

Joint universal lossy coding and identification of stationary mixing sources with general alphabets

We consider the problem of joint universal variable-rate lossy coding and identification for parametric classes of stationary $β$-mixing sources with general (Polish) alphabets. Compression performance is measured in terms of Lagrangians, while identification performance is measured by the variational distance between the true source and the estimated source. Provided that the sources are mixing at a sufficiently fast rate and satisfy certain smoothness and Vapnik-Chervonenkis learnability conditions, it is shown that, for bounded metric distortions, there exist universal schemes for joint lossy compression and identification whose Lagrangian redundancies converge to zero as $\sqrt{V_n \log n /n}$ as the block length $n$ tends to infinity, where $V_n$ is the Vapnik-Chervonenkis dimension of a certain class of decision regions defined by the $n$-dimensional marginal distributions of the sources; furthermore, for each $n$, the decoder can identify $n$-dimensional marginal of the active source up to a ball of radius $O(\sqrt{V_n\log n/n})$ in variational distance, eventually with probability one. The results are supplemented by several examples of parametric sources satisfying the regularity conditions.

cs.IT↗
arXiv2007

Joint fixed-rate universal lossy coding and identification of continuous-alphabet memoryless sources

The problem of joint universal source coding and identification is considered in the setting of fixed-rate lossy coding of continuous-alphabet memoryless sources. For a wide class of bounded distortion measures, it is shown that any compactly parametrized family of $\R^d$-valued i.i.d. sources with absolutely continuous distributions satisfying appropriate smoothness and Vapnik--Chervonenkis learnability conditions, admits a joint scheme for universal lossy block coding and parameter estimation, such that when the block length $n$ tends to infinity, the overhead per-letter rate and the distortion redundancies converge to zero as $O(n^{-1}\log n)$ and $O(\sqrt{n^{-1}\log n})$, respectively. Moreover, the active source can be determined at the decoder up to a ball of radius $O(\sqrt{n^{-1} \log n})$ in variational distance, asymptotically almost surely. The system has finite memory length equal to the block length, and can be thought of as blockwise application of a time-invariant nonlinear filter with initial conditions determined from the previous block. Comparisons are presented with several existing schemes for universal vector quantization, which do not include parameter estimation explicitly, and an extension to unbounded distortion measures is outlined. Finally, finite mixture classes and exponential families are given as explicit examples of parametric sources admitting joint universal compression and modeling schemes of the kind studied here.

cs.IT↗
arXiv2007

Joint universal lossy coding and identification of stationary mixing sources

The problem of joint universal source coding and modeling, treated in the context of lossless codes by Rissanen, was recently generalized to fixed-rate lossy coding of finitely parametrized continuous-alphabet i.i.d. sources. We extend these results to variable-rate lossy block coding of stationary ergodic sources and show that, for bounded metric distortion measures, any finitely parametrized family of stationary sources satisfying suitable mixing, smoothness and Vapnik-Chervonenkis learnability conditions admits universal schemes for joint lossy source coding and identification. We also give several explicit examples of parametric sources satisfying the regularity conditions.

cs.IT↗
arXiv2007

Learning from compressed observations

The problem of statistical learning is to construct a predictor of a random variable $Y$ as a function of a related random variable $X$ on the basis of an i.i.d. training sample from the joint distribution of $(X,Y)$. Allowable predictors are drawn from some specified class, and the goal is to approach asymptotically the performance (expected loss) of the best predictor in the class. We consider the setting in which one has perfect observation of the $X$-part of the sample, while the $Y$-part has to be communicated at some finite bit rate. The encoding of the $Y$-values is allowed to depend on the $X$-values. Under suitable regularity conditions on the admissible predictors, the underlying family of probability distributions and the loss function, we give an information-theoretic characterization of achievable predictor performance in terms of conditional distortion-rate functions. The ideas are illustrated on the example of nonparametric regression in Gaussian noise.

cs.IT↗
arXiv2006

Joint universal lossy coding and identification of i.i.d. vector sources

The problem of joint universal source coding and modeling, addressed by Rissanen in the context of lossless codes, is generalized to fixed-rate lossy coding of continuous-alphabet memoryless sources. We show that, for bounded distortion measures, any compactly parametrized family of i.i.d. real vector sources with absolutely continuous marginals (satisfying appropriate smoothness and Vapnik--Chervonenkis learnability conditions) admits a joint scheme for universal lossy block coding and parameter estimation, and give nonasymptotic estimates of convergence rates for distortion redundancies and variational distances between the active source and the estimated source. We also present explicit examples of parametric sources admitting such joint universal compression and modeling schemes.

cs.IT↗
arXiv2005

A complexity-regularized quantization approach to nonlinear dimensionality reduction

We consider the problem of nonlinear dimensionality reduction: given a training set of high-dimensional data whose ``intrinsic'' low dimension is assumed known, find a feature extraction map to low-dimensional space, a reconstruction map back to high-dimensional space, and a geometric description of the dimension-reduced data as a smooth manifold. We introduce a complexity-regularized quantization approach for fitting a Gaussian mixture model to the training set via a Lloyd algorithm. Complexity regularization controls the trade-off between adaptation to the local shape of the underlying manifold and global geometric consistency. The resulting mixture model is used to design the feature extraction and reconstruction maps and to define a Riemannian metric on the low-dimensional data. We also sketch a proof of consistency of our scheme for the purposes of estimating the unknown underlying pdf of high-dimensional data.

cs.IT↗
arXiv2005

Operational distance and fidelity for quantum channels

We define and study a fidelity criterion for quantum channels, which we term the minimax fidelity, through a noncommutative generalization of maximal Hellinger distance between two positive kernels in classical probability theory. Like other known fidelities for quantum channels, the minimax fidelity is well-defined for channels between finite-dimensional algebras, but it also applies to a certain class of channels between infinite-dimensional algebras (explicitly, those channels that possess an operator-valued Radon--Nikodym density with respect to the trace in the sense of Belavkin--Staszewski) and induces a metric on the set of quantum channels which is topologically equivalent to the CB-norm distance between channels, precisely in the same way as the Bures metric on the density operators associated with statistical states of quantum-mechanical systems, derived from the well-known fidelity (`generalized transition probability') of Uhlmann, is topologically equivalent to the trace-norm distance.

quant-ph↗
arXiv2003

A Phase Transition and Stochastic Domination in Pippenger's Probabilistic Failure Model for Boolean Networks with Unreliable Gates

We study Pippenger's model of Boolean networks with unreliable gates. In this model, the conditional probability that a particular gate fails, given the failure status of any subset of gates preceding it in the network, is bounded from above by some $ε$. We show that if we pick a Boolean network with $n$ gates at random according to the Barak-Erdős model of a random acyclic digraph, such that the expected edge density is $c n^{-1}\log n$, and if $ε$ is equal to a certain function of the size of the largest reflexive, transitive closure of a vertex (with respect to a particular realization of the random digraph), then Pippenger's model exhibits a phase transition at $c=1$. Namely, with probability $1-o(1)$ as $n\to\infty$, we have the following: for $0 \le c \le 1$, the minimum of the probability that no gate has failed, taken over all probability distributions of gate failures consistent with Pippenger's model, is equal to $o(1)$, whereas for $c >1$ it is equal to $\exp(-\frac{c}{e(c-1)}) + o(1)$. We also indicate how a more refined analysis of Pippenger's model, e.g., for the purpose of estimating probabilities of monotone events, can be carried out using the machinery of stochastic domination.

math.PR↗
arXiv2003

Radon-Nikodym derivatives of quantum operations

Given a completely positive (CP) map $T$, there is a theorem of the Radon-Nikodym type [W.B. Arveson, Acta Math. {\bf 123}, 141 (1969); V.P. Belavkin and P. Staszewski, Rep. Math. Phys. {\bf 24}, 49 (1986)] that completely characterizes all CP maps $S$ such that $T-S$ is also a CP map. This theorem is reviewed, and several alternative formulations are given along the way. We then use the Radon-Nikodym formalism to study the structure of order intervals of quantum operations, as well as a certain one-to-one correspondence between CP maps and positive operators, already fruitfully exploited in many quantum information-theoretic treatments. We also comment on how the Radon-Nikodym theorem can be used to derive norm estimates for differences of CP maps in general, and of quantum operations in particular.

math-ph↗
arXiv2003

Scaling and renormalization in fault-tolerant quantum computers

This work is concerned with phrasing the concepts of fault-tolerant quantum computation within the framework of disordered systems, Bernoulli site percolation in particular. We show how the so-called "threshold theorems" on the possibility of fault-tolerant quantum computation with constant error rate can be cast as a renormalization (coarse-graining) of the site percolation process describing the occurrence of errors during computation. We also use percolation techniques to derive a trade-off between the complexity overhead of the fault-tolerant circuit and the threshold error rate.

quant-ph↗
arXiv2003

Quantum system identification

We formulate and study, in general terms, the problem of quantum system identification, i.e., the determination (or estimation) of unknown quantum channels through their action on suitably chosen input density operators. We also present a quantitative analysis of the worst-case performance of these schemes.

quant-ph↗
arXiv2003

Dynamical Aspects of Information Storage in Quantum-Mechanical Systems

We study information storage in noisy quantum registers and computers using the methods of statistical dynamics. We develop the concept of a strictly contractive quantum channel in order to construct mathematical models of physically realizable, i.e., nonideal, quantum registers and computers. Strictly contractive channels are simple enough, yet exhibit very interesting features, which are meaningful from the physical point of view. In particular, they allow us to incorporate the crucial assumption of finite precision of all experimentally realizable operations. Strict contractivity also helps us gain insight into the thermodynamics of noisy quantum evolutions (approach to equilibrium). Our investigation into thermodynamics focuses on the entropy-energy balance in quantum registers and computers under the influence of strictly contractive noise. Using entropy-energy methods, we are able to appraise the thermodynamical resources needed to maintain reliable operation of the computer. We also obtain estimates of the largest tolerable error rate. Finally, we explore the possibility of going beyond the standard circuit model of error correction, namely constructing quantum memory devices on the basis of interacting particle systems at low temperatures.

quant-ph↗
arXiv2001

Strictly contractive quantum channels and physically realizable quantum computers

We study the robustness of quantum computers under the influence of errors modelled by strictly contractive channels. A channel $T$ is defined to be strictly contractive if, for any pair of density operators $ρ,σ$ in its domain, $\| Tρ- Tσ\|_1 \le k \| ρ-σ\|_1$ for some $0 \le k < 1$ (here $\| \cdot \|_1$ denotes the trace norm). In other words, strictly contractive channels render the states of the computer less distinguishable in the sense of quantum detection theory. Starting from the premise that all experimental procedures can be carried out with finite precision, we argue that there exists a physically meaningful connection between strictly contractive channels and errors in physically realizable quantum computers. We show that, in the absence of error correction, sensitivity of quantum memories and computers to strictly contractive errors grows exponentially with storage time and computation time respectively, and depends only on the constant $k$ and the measurement precision. We prove that strict contractivity rules out the possibility of perfect error correction, and give an argument that approximate error correction, which covers previous work on fault-tolerant quantum computation as a special case, is possible.

quant-ph↗
arXiv2001

A fidelity measure for quantum channels

We propose a fidelity measure for quantum channels in a straightforward analogy to the corresponding mixed-state fidelity of Jozsa. We describe properties of this fidelity measure and discuss some applications of it to quantum information science.

quant-ph↗
arXiv2001

Generation and manipulation of squeezed states of light in optical networks for quantum communication and computation

We analyze a fiber-optic component which could find multiple uses in novel information-processing systems utilizing squeezed states of light. Our approach is based on the phenomenon of photon-number squeezing of soliton noise after the soliton has propagated through a nonlinear optical fiber. Applications of this component in optical networks for quantum computation and quantum cryptography are discussed.

quant-ph↗