The theory of series cyclic queues is generalized to other configurations such as branching series queueing models. An equivalence relation is developed between time average steady state probabilities and the stationary probabilities of the imbedded Markov Chain of a series cyclic queue. A statistical inference model is developed to estimate the mean exponential service rate for a series cyclic queue model.

Topics: DTIC Archive, CALIFORNIA UNIV BERKELEY OPERATIONS RESEARCH CENTER, *QUEUEING THEORY, PERMUTATIONS,...

The research described in this report was conducted in fulfillment of Project MM-1572, A Complex Approach to UXO Discrimination: Combining Advanced EMI Forward and Statistical Signal Processing, submitted to the Strategic Environmental Research and Development Program (SERDP) in response to the Munitions Management Statement of Need MMSON-07-04, Advanced Technologies for Detection, Discrimination, and Remediation of Munitions and Explosives of Concern (MEC): UXO Technology. The well-known and...

Topics: DTIC Archive, SKY RESEARCH INC ASHLAND OR, *UNEXPLODED AMMUNITION, DISCRIMINATION, ELECTROMAGNETIC...

Topics: DTIC Archive, Price, James F, BOEING SCIENTIFIC RESEARCH LABS SEATTLE WA, *INTEGRALS, FUNCTIONAL...

Importance sampling is a technique that is commonly used to speed up Monte Carlo simulation of rare events. However, little is known regarding the design of efficient importance sampling algorithms in the context of queueing networks. The standard approach, which simulates the system using an a priori fixed change of measure suggested by large deviation analysis, has been shown to fail in even the simplest network setting (e.g., a two-node tandem network). Exploiting connections between...

Topics: DTIC Archive, Dupuis, Paul, BROWN UNIV PROVIDENCE RI DIV OF APPLIED MATHEMATICS, *STATISTICAL...

This report summarizes the progress made since November 1, 1984 under this grant. The primary contributions have been on nonlinear smoothers, quantization, and nonparametric and robust detection and filtering.

Topics: DTIC Archive, Kassam,S A, PENNSYLVANIA UNIV PHILADELPHIA, *SIGNAL PROCESSING, *DETECTION,...

Topics: DTIC Archive, BELLMAN, RICHARD, RAND CORP SANTA MONICA CA, *SCHEDULING, *STATISTICAL PROCESSES,...

This grant led to developments in flexible models for complex time series in a range of applications with a focus on Bayesian and Bayesian nonparametric methods. Three fundamental challenges were tackled: (i) capturing evolving correlations in high-dimensional time series with possible missing or irregularly-spaced observations, (ii) performing diverse subset selection over time, and (iii) automatically learning an unknown set of simple underlying temporal structures to describe complex...

Topics: DTIC Archive, WHARTON SCHOOL PHILADELPHIA PA, *STATISTICAL PROCESSES, CLASSIFICATION, MARKOV...

This report details the application of the SIG statistical learning approach to UXO discrimination for Camp Butner, North Carolina. This technology has been developed and validated under previous SERDP/ESTCP efforts by SIG and Duke University. Specific core technologies were used in this discrimination. These technologies fall broadly into the four analysis categories: the sensor/target model, feature selection, classification, and active label selection. The non-linear classifier outperformed...

Topics: DTIC Archive, SIGNAL INNOVATIONS GROUP INC DURHAM NC, *UNEXPLODED AMMUNITION, CLASSIFICATION,...

Sampling designs in common use: These are all probability-based (statistical) and all are well documented in the environmental literature: Simple Random, Systematic/Grid, Stratified, Composite, Incremental, Rank Set, Hot spot, and Adaptive Cluster.

Topics: DTIC Archive, NEPTUNE AND CO INC DENVER CO, *SAMPLING, STATISTICAL PROCESSES, WORKSHOPS

High spatial resolution observations of the H(alpha)-emitting wind structure associated with the luminous blue variable star P Cygni were obtained with the Navy Prototype Optical Interferometer. These observations represent the most comprehensive interferometric data set on P Cyg to date. We demonstrate how the apparent size of the H(alpha)-emitting region of the wind structure of P Cyg compares between the 2005, 2007, and 2008 observing seasons and how this relates to the H(alpha) line...

Topics: DTIC Archive, NAVAL OBSERVATORY FLAGSTAFF AZ, *VARIABLE STARS, *WIND, MATHEMATICAL MODELS, OPTICAL...

The report is prompted by the rash of good mathematical research on the problem of the moments of the number of crossings of a curve by a stochastic process. It is felt that this problem is important to some engineers and designers who may find a brief introduction to these recent publications useful.

Topics: DTIC Archive, Crawford, Gordon B, *STATISTICAL PROCESSES, EQUATIONS, MATHEMATICS, PROBABILITY,...

This dissertation presents new techniques for solving large sparse unsymmetric linear systems on high performance computers, using Gaussian elimination with partial pivoting. The efficiencies of the new algorithms are demonstrated for matrices from various fields and for a variety of high performance machines. In the first part we discuss optimizations of a sequential algorithm to exploit the memory hierarchies that exist in most RISC-based superscalar computers. Our key contribution is to...

Topics: DTIC Archive, CALIFORNIA UNIV BERKELEY GRADUATE DIV, *DISTRIBUTED COMPUTING, ALGORITHMS,...

Distributed multi-sensor fusion has been widely used in military and civilian applications. In the statistical sensor fusion domain, the design of an optimal fusion processor usually requires the joint statistics of the local sensor outputs. When accurate joint statistical knowledge is not readily available, popular solutions are either to estimate the joint statistics from training data or to simply assume independence of the data. Although it is well known that a fusion solution constructed...

Topics: DTIC Archive, TELEDYNE SCIENTIFIC AND IMAGING LLC DURHAM NC, *SENSOR FUSION, STATISTICAL PROCESSES

Both observational and laboratory studies suggest that human beings find it difficult to fabricate truly random numbers. Any sufficiently large set of nominally random number sequences, irrespective of its source, may be analyzed to determine if the criterion of randomness is satisfied. Statistical methods have, therefore, been applied forensically to detect anomalies in accounting data, scientific data sets, and voting data. As a case study, we apply methods developed by Beber and Scacco to...

Topics: DTIC Archive, NAVAL RESEARCH LAB WASHINGTON DC, *ELECTIONS, ANOMALIES, DETECTION, RANDOM VARIABLES,...

We introduce a novel generalization of the compound Gaussian (CG) (or Gaussian Scale Mixture) distribution which extends the Gaussian component of the CG model to a multilinear distribution. The resulting model, which we call the Multilinear Compound Gaussian (MCG) distribution, subsumes both GSM and the previously developed MICA distributions as complementary special cases; thereby allowing us to model a richer class of stochastic phenomena. First we derive the structural characterization of...

Topics: DTIC Archive, NAVAL RESEARCH LAB WASHINGTON DC RADAR DIV, *IMAGE PROCESSING, RADAR IMAGES,...

Topics: DTIC Archive, Alexander, M., AEROSPACE RESEARCH LABS WRIGHT-PATTERSON AFB OH, *ANALYSIS OF...

Our goals are to develop efficient means for statistically forecasting ocean currents and Lagrangian trajectories. Objectives are to formulate and test methods for predicting flows and trajectories that represent evolution of uncertainty and feedback of the uncertain component upon the mean. Such statistical predictions can be obtained via ensemble forecasts, but these are computationally expensive and suffer from statistical noise due to finite ensemble size. We seek statistical forecast...

Topics: DTIC Archive, INSTITUTE OF OCEAN SCIENCES SIDNEY (CANADA), *OCEAN CURRENTS, *STATISTICAL PROCESSES,...

A Zero-Knowledge PCP (ZK-PCP) is a randomized PCP such that the view of any (perhaps cheating) efficient verifier can be efficiently simulated up to small statistical distance. Kilian Petrank, and Tardos (STOC '97) constructed ZK-PCPs for all languages in NEXP. Ishai Mahmoody, and Sahai (TCC '12), motivated by cryptographic applications, revisited the possibility of efficient ZK-PCPs for all L is an element of NP where the PCP is encoded as a polynomial-size circuit that given a query i returns...

Topics: DTIC Archive, CORNELL UNIV ITHACA NY DEPT OF COMPUTER SCIENCE, *CRYPTOGRAPHY, ENTROPY, LANGUAGE,...

Topics: DTIC Archive, Carson, Daniel H, *INTELLIGIBILITY, *AUDITORY PERCEPTION, TEST METHODS,...

This work considers a computationally and statistically efficient parameter estimation method for a wide class of latent variable models--including Gaussian mixture models, hidden Markov models, and latent Dirichlet allocation--which exploits a certain tensor structure in their low-order observable moments (typically, of second- and third-order). Specifically, parameter estimation is reduced to the problem of extracting a certain (orthogonal) decomposition of a symmetric tensor derived from the...

Topics: DTIC Archive, CALIFORNIA UNIV IRVINE, *MATHEMATICAL MODELS, MARKOV PROCESSES, STATISTICAL...

We derive an exact p-value for testing a global null hypothesis in a general adaptive regression setting. Our approach uses the Kac-Rice formula (as described in Adler & Taylor 2007) applied to the problem of maximizing a Gaussian process. The resulting test statistic has a known distribution in finite samples, assuming Gaussian errors. We examine this test statistic in the case of the lasso, group lasso, principal components and matrix completion problems. For the lasso problem, our test...

Topics: DTIC Archive, STANFORD UNIV CA DEPT OF STATISTICS, *STATISTICAL PROCESSES, REGRESSION ANALYSIS

The formulae for estimating a mean will not work if the objective is the detection of the presence of a potential hot-spot . The problem becomes: 1.) How big is the potential hotspot; 2.) How large a mesh-spacing is needed; and 3.) What is the chance of missing a hotspot.

Topics: DTIC Archive, NEPTUNE AND CO INC DENVER CO, *SAMPLING, *STATISTICAL PROCESSES, GRIDS(COORDINATES),...

It is generally accepted that kinetic theory based on a statistical description of the gas provides a valid framework to describe processes in a rarefied regime or at small scales. Introductions into kinetic theory and its core, Boltzmann equation, can be found in many text books like Cercignani (1988), Cercignani (2000), Chapman and Cowling (1970) or Vincenti and Kruger (1965). The main variable used to describe the gas is the distribution function or probability density of the particle...

Topics: DTIC Archive, RWTH AACHEN UNIV (GERMANY), *KINETIC THEORY, BOLTZMANN EQUATION, GASES, GERMANY,...

This final technical report describes an effort to develop a comprehensive and accurate map of the Internet using passive measurements, diverse data sets and statistical learning methods. The effort passively collected a comprehensive set of Internet traffic and topology information. It captured a wide range of sources, including the Akamai data, Honeypot traffic, Boarder Gateway Protocol (BGP) traffic, background radiation, and distributed packet capture at Web servers and other network choke...

Topics: DTIC Archive, DUKE UNIV DURHAM NC, *INTERNET, *MAPPING, NETWORK TOPOLOGY, SOFTWARE TOOLS,...

The purpose of this analysis is to permit management to quickly and inexpensively evaluate a real time data processing system and to express a statistical confidence in the validity of their evaluation.

Topics: DTIC Archive, PRENDERGAST, DONN J, DAVID TAYLOR MODEL BASIN WASHINGTON DC, *DATA PROCESSING, *TIME,...

The use of statistical techniques to characterize composite materials systems and their use in complex electromagnetic environments is discussed. Examples of the calculation of uncertainties and distribution functions of wave statistics are given.

Topics: DTIC Archive, Arnaut, L. R., NATIONAL PHYSICAL LAB TEDDINGTON (UNITED KINGDOM), *COMPOSITE...

Using the total probability theorem, we propose a method to calculate the failure rate of a linear vibratory system with random parameters excited by stationary Gaussian processes. The response of such a system is non-stationary because of the randomness of the input parameters. A space-filling design, such as optimal symmetric Latin hypercube sampling or maximin, is first used to sample the input parameter space. For each design point, the output process is stationary and Gaussian. We present...

Topics: DTIC Archive, ARMY TANK AUTOMOTIVE RESEARCH DEVELOPMENT AND ENGINEERING CENTER WARREN MI, *LINEAR...

Web monitoring systems report any changes on the target web pages by revisiting them frequently. As they are operated under significant constrains such as network and computing, it is necessary to minimize revisits with minimal delay and maximum coverage. Various statistical scheduling methods were proposed to resolve this problem. However they are static and cannot easily cope with events in the real world. This paper proposes a new scheduling method that manages unpredictable events. MCRDR...

Topics: DTIC Archive, Kang, Byeong Ho, TASMANIA UNIV HOBART (AUSTRALIA), *INTERNET, *SCHEDULING,...

This report discusses statistical analyses of the spatial pattern of metallic anomalies, buried and on the ground surface, detected during airborne surveys above two former Air Force bombing ranges: the former Pueblo Precision Bombing Range Number 2 in Otero County, Colorado, and the Victorville Precision Bombing Range in San Bernardino County, California. The main purpose of the analyses is to determine whether statistical properties of anomaly spatial patterns can be used to delineate areas...

Topics: DTIC Archive, CARNEGIE-MELLON UNIV PITTSBURGH PA, *PATTERN RECOGNITION, *STATISTICAL PROCESSES,...

Unless the primary source of noise in one of those which are theoretically tractable, the statistical properties of the noise have to be inferred from samples. We have developed some criteria to aid in the decision whether a sample can reasonably be assumed to have come from a Gaussian noise with predetermined parameters.

Topics: DTIC Archive, RAND CORP SANTA MONICA CA, *NOISE, *SAMPLING, DISTRIBUTION THEORY, FUNCTIONAL...

This study summarizes the results of two series of demographic projections covering each of the 52 largest Standard Metropolitan Statistical Areas (SMSA) in the United States. These are trial projections, the first steps in a continuing effort to anticipate the probable sizes and characteristics of future metropolitan populations. For each SMSA, the 1960 population was divided into 60 components (age, by sex and color); on the basis of explicit and detailed assumptions as to rates of birth,...

Topics: DTIC Archive, RAND CORP SANTA MONICA CA, *POPULATION, *MATHEMATICAL PREDICTION, ECONOMICS,...

Topics: DTIC Archive, ARROW, KENNETH J, STANFORD UNIV CA APPLIED MATHEMATICS AND STATISTICS LABS, *COSTS,...

Bayes estimation of the arrival rate of a Poisson process is studied in this paper. For any loss function in the family L sub p, a simple sequential procedure Tau sub p is introduced which, based on the criterion of minimizing expected cost (estimation error plus sampling cost), is either optimal or asymptotically optimal. The procedure Tau sub p is compared to Type 1 and 2 censoring - the comparison should be useful to experimenters choosing between the three sampling plans.

Topics: DTIC Archive, Wardrop,Robert L, WISCONSIN UNIV-MADISON MATHEMATICS RESEARCH CENTER, *BAYES THEOREM,...

Percentiles and linear combinations of order statistics are statistics which are sometimes preferred to averages because they can be less sensitive to the presence of a few wild observations. It is well known that for large samples, both percentiles and linear combinations of order statistics resemble averages in that the appropriately normalized statistic is approximately normally distributed, with parameters which depend on the underlying distributions. This paper shows that percentiles and...

Topics: DTIC Archive, Leurgans,Sue, WISCONSIN UNIV-MADISON MATHEMATICS RESEARCH CENTER, *ORDER STATISTICS,...

We use process level large deviation analysis to obtain the rate function for a general family of occupancy problems. Our interest is the asymptotics of the empirical distributions of various quantities (such as the fraction of urns that contain a given number of balls). In the general setting, balls are allowed to land in a given urn depending on the urn's contents prior to the throw. We discuss a parametric family of statistical models which includes Maxwell-Boltzmann, Bose-Einstein and...

Topics: DTIC Archive, Zhang, Jim X, BROWN UNIV PROVIDENCE RI DIV OF APPLIED MATHEMATICS, *STATISTICAL...

The optimal statistical control of a simple production process which has only two underlying states, good and bad, is studied. The produced items are classified as good or defective, a cost being associated with each defective item produced. A cost is also charged for repairing the process, which has the effect of returning the process to the good state. Other than immediately after repair, the process state is assumed unknown. One seeks a statistical control rule which, based on the quality...

Topics: DTIC Archive, STANFORD UNIV CA DEPT OF STATISTICS, *OPERATIONS RESEARCH, *STATISTICAL PROCESSES,...

The purpose of the workshop was to explore the current state of research and applications in nonlinear time series analysis and to explore the common ground between the statistical and mathematical dynamical approach to nonlinear temporal systems. Attention was given to the practical use of theoretical developments in this topic. The Workshop had three main aspects -- Probabilistic aspects, Statistical aspects, and Dynamical systems. (edc)

Topics: DTIC Archive, EDINBURGH UNIV (UNITED KINGDOM), *NONLINEAR ANALYSIS, *TIME SERIES ANALYSIS,...

The only self-similar stable processes with stationary increments whose left-equivalent (resp. right-equivalent) stationary processes are nonanticipating (resp. fully anticipating) moving averages are the left (resp. right) linear fractional Levy motions.

Topics: DTIC Archive, Cambanis, Stamatis, NORTH CAROLINA UNIV AT CHAPEL HILL CENTER FOR STOCHASTIC...

A correspondence is obtained between the epochs of a Yule process and orders statistics from an exponential distribution. This correspondence is used to obtain some properties of the Yule process, which are in turn used to solve an optimatization problem. This problem involves the finding of an optimal evacuation time under a model in which individuals arrive to a contaminated area according to a Yule process with immigration.

Topics: DTIC Archive, Brown, Mark, GEORGE WASHINGTON UNIV WASHINGTON DC DEPT OF STATISTICS, *STATISTICAL...

An identity is generalized to an arbitrary number of variables. The proof of the generalized identity is obtained by computing transition probabilities in a Markov Chain.

Topics: DTIC Archive, MASAITIS, C, ARMY BALLISTIC RESEARCH LAB ABERDEEN PROVING GROUND MD, *ALGEBRA,...

The effort documented in this report includes an analysis of historical data related to government R&D program cost overruns and a preliminary assessment of the implications of such analysis for management reserves for related high-risk DARPA programs. This assessment will ultimately lead to the specification of a process to guide DARPA management in the establishment of budgetary reserves to account for technological risk and to periodically update projections of selected program cost...

Topics: DTIC Archive, MERIDIAN CORP FALLS CHURCH VA, *COST ESTIMATES, *COST OVERRUNS, DATA BASES, RISK,...

This report deals with an application of double sampling in the area of robustness. Configural polysampling is a technique which allows a detailed comparison of existing estimator and helps in finding small-sample-optimal estimators. The technique involves sampling across configurations. The associated sampling error can be reduced by using double sampling. Formulas for doing this are given and demonstrated in an example. (Author)

Topics: DTIC Archive, Morgenthaler,S, PRINCETON UNIV NJ DEPT OF STATISTICS, *SAMPLING, *STATISTICAL...

A robust version of Akaike's model selection procedure for regression models is introduced and its relationship with robust testing procedures is discussed. (Author)

Topics: DTIC Archive, Ronchetti,E, PRINCETON UNIV NJ DEPT OF STATISTICS, *MATHEMATICAL MODELS, *REGRESSION...

In order to understand the problems intractable problems nature presents us, we are forced to make simplifications till we arrive at simple analytical models that we are capable of understanding. Such canonical models are rare, but useful as tools for constructing more realistic models that we can use analyze nature. The class of analytical models for clutter analysis limited to those that consist of various amplitude models with the phase noise assumed to have a probability density function...

Topics: DTIC Archive, NAVAL SURFACE WARFARE CENTER DAHLGREN DIV VA, *RADAR CLUTTER, *PROBABILITY DENSITY...

Hyperspectral data has become a critical tool for use by military analysts and planners. The capture of fine spectral information enables the generation of information products which could not be produced using traditional imaging means. The challenge facing hyperspectral technology, as an operational capability, is with conversion of the raw sensor data into a useful information product that is accurate and reliable. Traditional approaches for processing hyperspectral data have largely focused...

Topics: DTIC Archive, ROCHESTER INST OF TECH NY, *HYPERSPECTRAL IMAGERY, ALGORITHMS, IMAGE PROCESSING,...

This note gives a justification for the interchange of limiting processes required in Doob's 'heuristic' approach to the Kolmogorov limiting distribution of the maximum deviation between a theoretical and an empirical distribution function.

Topics: DTIC Archive, RAND CORP SANTA MONICA CA, *STATISTICAL FUNCTIONS, *STATISTICAL PROCESSES,...

General additive functions called rewards are defined on a 'regular' finite-state Markov-renewal process. The asymptotic form of the mean total reward in (O, t) has previously been obtained, and it is known that the total rewards are joint-normally distributed as t approaches infinity. This paper finds the dominant asymptotic term in the covariance of the total rewards as a simple function of the moments of the per-transition rewards, and the 'bias' term of the mean total rewards. Special...

Topics: DTIC Archive, CALIFORNIA UNIV BERKELEY OPERATIONS RESEARCH CENTER, *STATISTICAL PROCESSES, ANALYSIS...

Total Quality Management (TQM) is a DoD initiative for continuously improving performance at every level, in every area of DoD responsibility. Implementing this philosophy will require a cultured change in th defense community. It will also require the intelligent use of appropriate analysis techniques. This report describes the basic analytical tools used in TQM: process flow charts, Ishikawa charts, Statistical process control, Histograms, Pareto diagrams, Scattergrams and the Shewhart cycle....

Topics: DTIC Archive, Coppola, Anthony, ROME LAB ROME NY, *QUALITY, MANAGEMENT, COMPREHENSION, MILITARY...

Contents: Statistical Inference in Markov-Renewal Processes; A Note on Deficit, Excess, and Spread in a Markov-Renewal Process; Near-Parallel Constrained Objectives in Integer Programs; Programming Under Nonlinear Constraints by Unconstrained Minimization: A Primal-Dual Method; Le Methods du Chemin CritiqueApplication aux Programmes de Production et d'Etudes de la Methode P.E.R.T. et ses Variantes; Les Problems d'Ordonnencement-Application et Methodes; Scientific Decision Making in Business:...

Topics: DTIC Archive, CALIFORNIA UNIV BERKELEY OPERATIONS RESEARCH CENTER, *OPERATIONS RESEARCH, ABSTRACTS,...

We report on experiments for the Related Entity Finding task in which we focus on only using Wikipedia as a target corpus in which to identify (related) entitities. Our approach is based on co-occurrences between the source entity and potential target entities. We observe improvements in performance when a context-independent co-occurrence model is combined with context-dependent co-occurrence models in which we stress the importance of the expected relation between source and target entity....

Topics: DTIC Archive, AMSTERDAM UNIV (NETHERLANDS), *TEXT PROCESSING, *STATISTICAL PROCESSES, *RANKING,...