Results 1  10
of
34,420
Graphical models, exponential families, and variational inference
, 2008
"... The formalism of probabilistic graphical models provides a unifying framework for capturing complex dependencies among random variables, and building largescale multivariate statistical models. Graphical models have become a focus of research in many statistical, computational and mathematical fiel ..."
Abstract

Cited by 800 (26 self)
 Add to MetaCart
The formalism of probabilistic graphical models provides a unifying framework for capturing complex dependencies among random variables, and building largescale multivariate statistical models. Graphical models have become a focus of research in many statistical, computational and mathematical fields, including bioinformatics, communication theory, statistical physics, combinatorial optimization, signal and image processing, information retrieval and statistical machine learning. Many problems that arise in specific instances — including the key problems of computing marginals and modes of probability distributions — are best studied in the general setting. Working with exponential family representations, and exploiting the conjugate duality between the cumulant function and the entropy for exponential families, we develop general variational representations of the problems of computing likelihoods, marginal probabilities and most probable configurations. We describe how a wide varietyof algorithms — among them sumproduct, cluster variational methods, expectationpropagation, mean field methods, maxproduct and linear programming relaxation, as well as conic programming relaxations — can all be understood in terms of exact or approximate forms of these variational representations. The variational approach provides a complementary alternative to Markov chain Monte Carlo as a general source of approximation methods for inference in largescale statistical models.
A Parametric Texture Model based on Joint Statistics of Complex Wavelet Coefficients
 INTERNATIONAL JOURNAL OF COMPUTER VISION
, 2000
"... We present a universal statistical model for texture images in the context of an overcomplete complex wavelet transform. The model is parameterized by a set of statistics computed on pairs of coefficients corresponding to basis functions at adjacent spatial locations, orientations, and scales. We de ..."
Abstract

Cited by 409 (13 self)
 Add to MetaCart
We present a universal statistical model for texture images in the context of an overcomplete complex wavelet transform. The model is parameterized by a set of statistics computed on pairs of coefficients corresponding to basis functions at adjacent spatial locations, orientations, and scales. We develop an efficient algorithm for synthesizing random images subject to these constraints, by iteratively projecting onto the set of images satisfying each constraint, and we use this to test the perceptual validity of the model. In particular, we demonstrate the necessity of subgroups of the parameter set by showing examples of texture synthesis that fail when those parameters are removed from the set. We also demonstrate the power of our model by successfully synthesizing examples drawn from a diverse collection of artificial and natural textures.
Endtoend available bandwidth: Measurement methodology, dynamics, and relation with TCP throughput
 In Proceedings of ACM SIGCOMM
, 2002
"... The available bandwidth (availbw) in a network path is of major importance in congestion control, streaming applications, QoS verification, server selection, and overlay networks. We describe an endtoend methodology, called SelfLoading Periodic Streams (SLoPS), for measuring availbw. The basic ..."
Abstract

Cited by 408 (20 self)
 Add to MetaCart
The available bandwidth (availbw) in a network path is of major importance in congestion control, streaming applications, QoS verification, server selection, and overlay networks. We describe an endtoend methodology, called SelfLoading Periodic Streams (SLoPS), for measuring availbw. The basic idea in SLoPS is that the oneway delays of a periodic packet stream show an increasing trend when the stream’s rate is higher than the availbw. We implemented SLoPS in a tool called pathload. The accuracy of the tool has been evaluated with both simulations and experiments over realworld Internet paths. Pathload is nonintrusive, meaning that it does not cause significant increases in the network utilization, delays, or losses. We used pathload to evaluate the variability (‘dynamics’) of the availbw in some paths that cross USA and Europe. The availbw becomes significantly more variable in heavily utilized paths, as well as in paths with limited capacity (probably due to a lower degree of statistical multiplexing). We finally examine the relation between availbw and TCP throughput. A persistent TCP connection can be used to roughly measure the availbw in a path, but TCP saturates the path, and increases significantly the path delays and jitter.
Securitycontrol methods for statistical databases: a comparative study
 ACM Computing Surveys
, 1989
"... This paper considers the problem of providing security to statistical databases against disclosure of confidential information. Securitycontrol methods suggested in the literature are classified into four general approaches: conceptual, query restriction, data perturbation, and output perturbation. ..."
Abstract

Cited by 405 (0 self)
 Add to MetaCart
This paper considers the problem of providing security to statistical databases against disclosure of confidential information. Securitycontrol methods suggested in the literature are classified into four general approaches: conceptual, query restriction, data perturbation, and output perturbation. Criteria for evaluating the performance of the various securitycontrol methods are identified. Securitycontrol methods that are based on each of the four approaches are discussed, together with their performance with respect to the identified evaluation criteria. A detailed comparative analysis of the most promising methods for protecting dynamiconline statistical databases is also presented. To date no single securitycontrol method prevents both exact and partial disclosures. There are, however, a few perturbationbased methods that prevent exact disclosure and enable the database administrator to exercise “statistical disclosure control. ” Some of these methods, however introduce bias into query responses or suffer from the O/l querysetsize problem (i.e., partial disclosure is possible in case of null query set or a query set of size 1). We recommend directing future research efforts toward developing new methods that prevent exact disclosure and provide statisticaldisclosure control, while at the same time do not suffer from the bias problem and the O/l querysetsize problem. Furthermore, efforts directed toward developing a biascorrection mechanism and solving the general problem of small querysetsize would help salvage a few of the current perturbationbased methods.
Reachability Analysis of Pushdown Automata: Application to ModelChecking
, 1997
"... We apply the symbolic analysis principle to pushdown systems. We represent (possibly infinite) sets of configurations of such systems by means of finitestate automata. In order to reason in a uniform way about analysis problems involving both existential and universal path quantification (like mode ..."
Abstract

Cited by 385 (39 self)
 Add to MetaCart
We apply the symbolic analysis principle to pushdown systems. We represent (possibly infinite) sets of configurations of such systems by means of finitestate automata. In order to reason in a uniform way about analysis problems involving both existential and universal path quantification (like modelchecking for branchingtime logics), we consider the more general class of alternating pushdown systems and use alternating finitestate automata as a representation structure for their sets of configurations. We give a simple and natural procedure to compute sets of predecessors for this representation structure. We apply this procedure and the automatatheoretic approach to modelchecking to define new modelchecking algorithms for pushdown systems and both linear and branchingtime properties. From these results we derive upper bounds for several modelchecking problems, and we also provide matching lower bounds, using reductions based on some techniques introduced by Walukiewicz.
ControlFlow Analysis of HigherOrder Languages
, 1991
"... representing the official policies, either expressed or implied, of ONR or the U.S. Government. Keywords: dataflow analysis, Scheme, LISP, ML, CPS, type recovery, higherorder functions, functional programming, optimising compilers, denotational semantics, nonstandard Programs written in powerful, ..."
Abstract

Cited by 362 (10 self)
 Add to MetaCart
representing the official policies, either expressed or implied, of ONR or the U.S. Government. Keywords: dataflow analysis, Scheme, LISP, ML, CPS, type recovery, higherorder functions, functional programming, optimising compilers, denotational semantics, nonstandard Programs written in powerful, higherorder languages like Scheme, ML, and Common Lisp should run as fast as their FORTRAN and C counterparts. They should, but they don’t. A major reason is the level of optimisation applied to these two classes of languages. Many FORTRAN and C compilers employ an arsenal of sophisticated global optimisations that depend upon dataflow analysis: commonsubexpression elimination, loopinvariant detection, inductionvariable elimination, and many, many more. Compilers for higherorder languages do not provide these optimisations. Without them, Scheme, LISP and ML compilers are doomed to produce code that runs slower than their FORTRAN and C counterparts. The problem is the lack of an explicit controlflow graph at compile time, something which traditional dataflow analysis techniques require. In this dissertation, I present a technique for recovering the controlflow graph of a Scheme program at compile time. I give examples of how this information can be used to perform several dataflow analysis optimisations, including copy propagation, inductionvariable elimination, uselessvariable elimination, and type recovery. The analysis is defined in terms of a nonstandard semantic interpretation. The denotational semantics is carefully developed, and several theorems establishing the correctness of the semantics and the implementing algorithms are proven. iii ivTo my parents, Julia and Olin. v viContents
Face recognition by independent component analysis
 IEEE Transactions on Neural Networks
, 2002
"... Abstract—A number of current face recognition algorithms use face representations found by unsupervised statistical methods. Typically these methods find a set of basis images and represent faces as a linear combination of those images. Principal component analysis (PCA) is a popular example of such ..."
Abstract

Cited by 333 (5 self)
 Add to MetaCart
Abstract—A number of current face recognition algorithms use face representations found by unsupervised statistical methods. Typically these methods find a set of basis images and represent faces as a linear combination of those images. Principal component analysis (PCA) is a popular example of such methods. The basis images found by PCA depend only on pairwise relationships between pixels in the image database. In a task such as face recognition, in which important information may be contained in the highorder relationships among pixels, it seems reasonable to expect that better basis images may be found by methods sensitive to these highorder statistics. Independent component analysis (ICA), a generalization of PCA, is one such method. We used a version of ICA derived from the principle of optimal information transfer through sigmoidal neurons. ICA was performed on face images in the FERET database under two different architectures, one which treated the images as random variables and the pixels as outcomes, and a second which treated the pixels as random variables and the images as outcomes. The first architecture found spatially local basis images for the faces. The second architecture produced a factorial face code. Both ICA representations were superior to representations based on PCA for recognizing faces across days and changes in expression. A classifier that combined the two ICA representations gave the best performance. Index Terms—Eigenfaces, face recognition, independent component analysis (ICA), principal component analysis (PCA), unsupervised learning. I.
Results 1  10
of
34,420