In this paper, we propose a hybrid method that uses stochastic and deterministic search to compute the maximum likelihood estimator of a low-rank count tensor with Poisson loss via state-of-theart local methods. Our approach is inspired by Simulated Annealing for global optimization and allows for fine-grain parameter tuning as well as adaptive updates to algorithm parameters. We present numerical results that indicate our hybrid approach can compute better approximations to the maximum likelihood estimator with less computation than the state-of-the-art methods by themselves.
We propose a novel statistical inference paradigm for zero-inflated multiway count data that dispenses with the need to distinguish between true and false zero counts. Our approach ignores all zero entries and applies zero-truncated Poisson regression on the positive counts. Inference is accomplished via tensor completion that imposes low-rank structure on the Poisson parameter space. Our main result shows that an $\textit{N}$-way rank-R parametric tensor 𝓜 ϵ (0, ∞)$I$Χ∙∙∙Χ$I$ generating Poisson observations can be accurately estimated from approximately $IR^2 \text{log}^2_2(I)$ non-zero counts for a nonnegative canonical polyadic decomposition. Several numerical experiments are presented demonstrating that our zero-truncated paradigm is comparable to the ideal scenario where the locations of false zero counts are known $\textit{a priori}$.
We present a novel approach to information retrieval and document analysis based on graph analytic methods. Traditional information retrieval methods use a set of terms to define a query that is applied against a document corpus to identify the documents most related to those terms. In contrast, we define a query as a set of documents of interest and apply the query by computing mean hitting times between this set and all other documents on a document similarity graph abstraction of the semantic relationships between all pairs of documents. We present the steps of our approach along with a simple example application illustrating how this approach can be used to find documents related to two or more documents or topics of interest.
Both the data science and scientific computing communities are embracing GPU acceleration for their most demanding workloads. For scientific computing applications, the massive volume of code and diversity of hardware platforms at supercomputing centers has motivated a strong effort toward performance portability. This property of a program, denoting its ability to perform well on multiple architectures and varied datasets, is heavily dependent on the choice of parallel programming model and which features of the programming model are used. In this paper, we evaluate performance portability in the context of a data science workload in contrast to a scientific computing workload, evaluating the same sparse matrix kernel on both. Among our implementations of the kernel in different performance-portable programming models, we find that many struggle to consistently achieve performance improvements using the GPU compared to simple one-line OpenMP parallelization on high-end multicore CPUs. We show one that does, and its performance approaches and sometimes even matches that of vendor-provided GPU math libraries.
Poisson Tensor Factorization (PTF) is an important data analysis method for analyzing patterns and relationships in multiway count data. In this work, we consider several algorithms for computing a low-rank PTF of tensors with sparse count data values via maximum likelihood estimation. Such an approach reduces to solving a nonlinear, non-convex optimization problem, which can leverage considerable parallel computation due to the structure of the problem. However, since the maximum likelihood estimator corresponds to the global minimizer of this optimization problem, it is important to consider how effective methods are at both leveraging this inherent parallelism as well as computing a good approximation to the global minimizer. In this work we present comparisons of multiple methods for PTF that illustrate the tradeoffs in computational efficiency and accurately computing the maximum likelihood estimator. We present results using synthetic and real-world data tensors to demonstrate some of the challenges when choosing a method for a given tensor.
Canonical Polyadic tensor decomposition using alternate Poisson regression (CP-APR) is an effective analysis tool for large sparse count datasets. One of the variants using projected damped Newton optimization for row subproblems (PDNR) offers quadratic convergence and is amenable to parallelization. Despite its potential effectiveness, PDNR performance on modern high performance computing (HPC) systems is not well understood. To remedy this, we have developed a parallel implementation of PDNR using Kokkos, a performance portable parallel programming framework supporting efficient runtime of a single code base on multiple HPC systems. We demonstrate that the performance of parallel PDNR can be poor if load imbalance associated with the irregular distribution of nonzero entries in the tensor data is not addressed. Preliminary results using tensors from the FROSTT data set indicate that using multiple kernels to address this imbalance when solving the PDNR row subproblems in parallel can improve performance, with up to 80% speedup on CPUs and 10-fold speedup on NVIDIA GPUs.
Tensor decomposition models play an increasingly important role in modern data science applications. One problem of particular interest is fitting a low-rank Canonical Polyadic (CP) tensor decomposition model when the tensor has sparse structure and the tensor elements are nonnegative count data. SparTen is a high-performance C++ library which computes a low-rank decomposition using different solvers: a first-order quasi-Newton or a second-order damped Newton method, along with the appropriate choice of runtime parameters. Since default parameters in SparTen are tuned to experimental results in prior published work on a single real-world dataset conducted using MATLAB implementations of these methods, it remains unclear if the parameter defaults in SparTen are appropriate for general tensor data. Furthermore, it is unknown how sensitive algorithm convergence is to changes in the input parameter values. This report addresses these unresolved issues with large-scale experimentation on three benchmark tensor data sets. Experiments were conducted on several different CPU architectures and replicated with many initial states to establish generalized profiles of algorithm convergence behavior.
Driven by the importance of relational aspects of data to decision-making, graph algorithms have been developed, based on simplified pairwise relationships, to solve a variety of problems. However, evidence has shown that hypergraphs - generalizations of graphs with (hyper)edges that connect any number of vertices - can better model complex, non-pairwise relationships in data and lead to better informed decisions. In this work, we compare graph and hypergraph models in the context of spectral clustering. For these problems, we demonstrate that hypergraphs are computationally more efficient and can better model complex, non-pairwise relationships for many datasets.
Cyber defense is an asymmetric battle today. We need to understand better what options are available for providing defenders with possible advantages. Our project combines machine learning, optimization, and game theory to obscure our defensive posture from the information the adversaries are able to observe. The main conceptual contribution of this research is to separate the problem of prediction, for which machine learning is used, and the problem of computing optimal operational decisions based on such predictions, coup led with a model of adversarial response. This research includes modeling of the attacker and defender, formulation of useful optimization models for studying adversarial interactions, and user studies to meas ure the impact of the modeling approaches in re alistic settings.
This reports describes extensions of DEDICOM (DEcomposition into DIrectional COMponents) data models [3] that incorporate bound and linear constraints. The main purpose of these extensions is to investigate the use of improved data models for unsupervised part-of-speech tagging, as described by Chew et al. [2]. In that work, a single domain, two-way DEDICOM model was computed on a matrix of bigram fre- quencies of tokens in a corpus and used to identify parts-of-speech as an unsupervised approach to that problem. An open problem identi ed in that work was the com- putation of a DEDICOM model that more closely resembled the matrices used in a Hidden Markov Model (HMM), speci cally through post-processing of the DEDICOM factor matrices. The work reported here consists of the description of several models that aim to provide a direct solution to that problem and a way to t those models. The approach taken here is to incorporate the model requirements as bound and lin- ear constrains into the DEDICOM model directly and solve the data tting problem as a constrained optimization problem. This is in contrast to the typical approaches in the literature, where the DEDICOM model is t using unconstrained optimization approaches, and model requirements are satis ed as a post-processing step.
Early 2010 saw a signi cant change in adversarial techniques aimed at network intrusion: a shift from malware delivered via email attachments toward the use of hidden, embedded hyperlinks to initiate sequences of downloads and interactions with web sites and network servers containing malicious software. Enterprise security groups were well poised and experienced in defending the former attacks, but the new types of attacks were larger in number, more challenging to detect, dynamic in nature, and required the development of new technologies and analytic capabilities. The Hybrid LDRD project was aimed at delivering new capabilities in large-scale data modeling and analysis to enterprise security operators and analysts and understanding the challenges of detection and prevention of emerging cybersecurity threats. Leveraging previous LDRD research e orts and capabilities in large-scale relational data analysis, large-scale discrete data analysis and visualization, and streaming data analysis, new modeling and analysis capabilities were quickly brought to bear on the problems in email phishing and spear phishing attacks in the Sandia enterprise security operational groups at the onset of the Hybrid project. As part of this project, a software development and deployment framework was created within the security analyst work ow tool sets to facilitate the delivery and testing of new capabilities as they became available, and machine learning algorithms were developed to address the challenge of dynamic threats. Furthermore, researchers from the Hybrid project were embedded in the security analyst groups for almost a full year, engaged in daily operational activities and routines, creating an atmosphere of trust and collaboration between the researchers and security personnel. The Hybrid project has altered the way that research ideas can be incorporated into the production environments of Sandias enterprise security groups, reducing time to deployment from months and years to hours and days for the application of new modeling and analysis capabilities to emerging threats. The development and deployment framework has been generalized into the Hybrid Framework and incor- porated into several LDRD, WFO, and DOE/CSL projects and proposals. And most importantly, the Hybrid project has provided Sandia security analysts with new, scalable, extensible analytic capabilities that have resulted in alerts not detectable using their previous work ow tool sets.
This SAND report summarizes the activities and outcomes of the Network and Ensemble Enabled Entity Extraction in Information Text (NEEEEIT) LDRD project, which addressed improving the accuracy of conditional random fields for named entity recognition through the use of ensemble methods.
Recent work on eigenvalues and eigenvectors for tensors of order m {>=} 3 has been motivated by applications in blind source separation, magnetic resonance imaging, molecular conformation, and more. In this paper, we consider methods for computing real symmetric-tensor eigenpairs of the form Ax{sup m-1} = {lambda}x subject to {parallel}x{parallel} = 1, which is closely related to optimal rank-1 approximation of a symmetric tensor. Our contribution is a novel shifted symmetric higher-order power method (SS-HOPM), which we showis guaranteed to converge to a tensor eigenpair. SS-HOPM can be viewed as a generalization of the power iteration method for matrices or of the symmetric higher-order power method. Additionally, using fixed point analysis, we can characterize exactly which eigenpairs can and cannot be found by the method. Numerical examples are presented, including examples from an extension of the method to fnding complex eigenpairs.
This report is a summary of the accomplishments of the 'Scalable Solutions for Processing and Searching Very Large Document Collections' LDRD, which ran from FY08 through FY10. Our goal was to investigate scalable text analysis; specifically, methods for information retrieval and visualization that could scale to extremely large document collections. Towards that end, we designed, implemented, and demonstrated a scalable framework for text analysis - ParaText - as a major project deliverable. Further, we demonstrated the benefits of using visual analysis in text analysis algorithm development, improved performance of heterogeneous ensemble models in data classification problems, and the advantages of information theoretic methods in user analysis and interpretation in cross language information retrieval. The project involved 5 members of the technical staff and 3 summer interns (including one who worked two summers). It resulted in a total of 14 publications, 3 new software libraries (2 open source and 1 internal to Sandia), several new end-user software applications, and over 20 presentations. Several follow-on projects have already begun or will start in FY11, with additional projects currently in proposal.
This report is a summary of the accomplishments of the 'Leveraging Multi-way Linkages on Heterogeneous Data' which ran from FY08 through FY10. The goal was to investigate scalable and robust methods for multi-way data analysis. We developed a new optimization-based method called CPOPT for fitting a particular type of tensor factorization to data; CPOPT was compared against existing methods and found to be more accurate than any faster method and faster than any equally accurate method. We extended this method to computing tensor factorizations for problems with incomplete data; our results show that you can recover scientifically meaningfully factorizations with large amounts of missing data (50% or more). The project has involved 5 members of the technical staff, 2 postdocs, and 1 summer intern. It has resulted in a total of 13 publications, 2 software releases, and over 30 presentations. Several follow-on projects have already begun, with more potential projects in development.
The problem of incomplete data - i.e., data with missing or unknown values - in multi-way arrays is ubiquitous in biomedical signal processing, network traffic analysis, bibliometrics, social network analysis, chemometrics, computer vision, communication networks, etc. We consider the problem of how to factorize data sets with missing values with the goal of capturing the underlying latent structure of the data and possibly reconstructing missing values (i.e., tensor completion). We focus on one of the most well-known tensor factorizations that captures multi-linear structure, CANDECOMP/PARAFAC (CP). In the presence of missing data, CP can be formulated as a weighted least squares problem that models only the known entries. We develop an algorithm called CP-WOPT (CP Weighted OPTimization) that uses a first-order optimization approach to solve the weighted least squares problem. Based on extensive numerical experiments, our algorithm is shown to successfully factorize tensors with noise and up to 99% missing data. A unique aspect of our approach is that it scales to sparse large-scale data, e.g., 1000 x 1000 x 1000 with five million known entries (0.5% dense). We further demonstrate the usefulness of CP-WOPT on two real-world applications: a novel EEG (electroencephalogram) application where missing data is frequently encountered due to disconnections of electrodes and the problem of modeling computer network traffic where data may be absent due to the expense of the data collection process.
Automated analysis of unstructured text documents (e.g., web pages, newswire articles, research publications, business reports) is a key capability for solving important problems in areas including decision making, risk assessment, social network analysis, intelligence analysis, scholarly research and others. However, as data sizes continue to grow in these areas, scalable processing, modeling, and semantic analysis of text collections becomes essential. In this paper, we present the ParaText text analysis engine, a distributed memory software framework for processing, modeling, and analyzing collections of unstructured text documents. Results on several document collections using hundreds of processors are presented to illustrate the exibility, extensibility, and scalability of the the entire process of text modeling from raw data ingestion to application analysis.
The aim of this project is to develop low dimension parametric (deterministic) models of complex networks, to use compressive sensing (CS) and multiscale analysis to do so and to exploit the structure of complex networks (some are self-similar under coarsening). CS provides a new way of sampling and reconstructing networks. The approach is based on multiresolution decomposition of the adjacency matrix and its efficient sampling. It requires preprocessing of the adjacency matrix to make it 'blocky' which is the biggest (combinatorial) algorithm challenge. Current CS reconstruction algorithm makes no use of the structure of a graph, its very general (and so not very efficient/customized). Other model-based CS techniques exist, but not yet adapted to networks. Obvious starting point for future work is to increase the efficiency of reconstruction.
The data in many disciplines such as social networks, web analysis, etc. is link-based, and the link structure can be exploited for many different data mining tasks. In this paper, we consider the problem of temporal link prediction: Given link data for time periods 1 through T, can we predict the links in time period T + 1? Specifically, we look at bipartite graphs changing over time and consider matrix- and tensor-based methods for predicting links. We present a weight-based method for collapsing multi-year data into a single matrix. We show how the well-known Katz method for link prediction can be extended to bipartite graphs and, moreover, approximated in a scalable way using a truncated singular value decomposition. Using a CANDECOMP/PARAFAC tensor decomposition of the data, we illustrate the usefulness of exploiting the natural three-dimensional structure of temporal link data. Through several numerical experiments, we demonstrate that both matrix- and tensor-based techniques are effective for temporal link prediction despite the inherent difficulty of the problem.
Automated processing, modeling, and analysis of unstructured text (news documents, web content, journal articles, etc.) is a key task in many data analysis and decision making applications. As data sizes grow, scalability is essential for deep analysis. In many cases, documents are modeled as term or feature vectors and latent semantic analysis (LSA) is used to model latent, or hidden, relationships between documents and terms appearing in those documents. LSA supplies conceptual organization and analysis of document collections by modeling high-dimension feature vectors in many fewer dimensions. While past work on the scalability of LSA modeling has focused on the SVD, the goal of our work is to investigate the use of distributed memory architectures for the entire text analysis process, from data ingestion to semantic modeling and analysis. ParaText is a set of software components for distributed processing, modeling, and analysis of unstructured text. The ParaText source code is available under a BSD license, as an integral part of the Titan toolkit. ParaText components are chained-together into data-parallel pipelines that are replicated across processes on distributed-memory architectures. Individual components can be replaced or rewired to explore different computational strategies and implement new functionality. ParaText functionality can be embedded in applications on any platform using the native C++ API, Python, or Java. The ParaText MPI Process provides a 'generic' text analysis pipeline in a command-line executable that can be used for many serial and parallel analysis tasks. ParaText can also be deployed as a web service accessible via a RESTful (HTTP) API. In the web service configuration, any client can access the functionality provided by ParaText using commodity protocols ... from standard web browsers to custom clients written in any language.
The problem of missing data is ubiquitous in domains such as biomedical signal processing, network traffic analysis, bibliometrics, social network analysis, chemometrics, computer vision, and communication networks|all domains in which data collection is subject to occasional errors. Moreover, these data sets can be quite large and have more than two axes of variation, e.g., sender, receiver, time. Many applications in those domains aim to capture the underlying latent structure of the data; in other words, they need to factorize data sets with missing entries. If we cannot address the problem of missing data, many important data sets will be discarded or improperly analyzed. Therefore, we need a robust and scalable approach for factorizing multi-way arrays (i.e., tensors) in the presence of missing data. We focus on one of the most well-known tensor factorizations, CANDECOMP/PARAFAC (CP), and formulate the CP model as a weighted least squares problem that models only the known entries. We develop an algorithm called CP-WOPT (CP Weighted OPTimization) using a first-order optimization approach to solve the weighted least squares problem. Based on extensive numerical experiments, our algorithm is shown to successfully factor tensors with noise and up to 70% missing data. Moreover, our approach is significantly faster than the leading alternative and scales to larger problems. To show the real-world usefulness of CP-WOPT, we illustrate its applicability on a novel EEG (electroencephalogram) application where missing data is frequently encountered due to disconnections of electrodes.
We present Poblano v1.0, a Matlab toolbox for solving gradient-based unconstrained optimization problems. Poblano implements three optimization methods (nonlinear conjugate gradients, limited-memory BFGS, and truncated Newton) that require only first order derivative information. In this paper, we describe the Poblano methods, provide numerous examples on how to use Poblano, and present results of Poblano used in solving problems from a standard test collection of unconstrained optimization problems.
Tensor decompositions are higher-order analogues of matrix decompositions and have proven to be powerful tools for data analysis. In particular, we are interested in the canonical tensor decomposition, otherwise known as the CANDECOMP/PARAFAC decomposition (CPD), which expresses a tensor as the sum of component rank-one tensors and is used in a multitude of applications such as chemometrics, signal processing, neuroscience, and web analysis. The task of computing the CPD, however, can be difficult. The typical approach is based on alternating least squares (ALS) optimization, which can be remarkably fast but is not very accurate. Previously, nonlinear least squares (NLS) methods have also been recommended; existing NLS methods are accurate but slow. In this paper, we propose the use of gradient-based optimization methods. We discuss the mathematical calculation of the derivatives and further show that they can be computed efficiently, at the same cost as one iteration of ALS. Computational experiments demonstrate that the gradient-based optimization methods are much more accurate than ALS and orders of magnitude faster than NLS.
Link analysis typically focuses on a single type of connection, e.g., two journal papers are linked because they are written by the same author. However, often we want to analyze data that has multiple linkages between objects, e.g., two papers may have the same keywords and one may cite the other. The goal of this paper is to show that multilinear algebra provides a tool for multilink analysis. We analyze five years of publication data from journals published by the Society for Industrial and Applied Mathematics. We explore how papers can be grouped in the context of multiple link types using a tensor to represent all the links between them. A PARAFAC decomposition on the resulting tensor yields information similar to the SVD decomposition of a standard adjacency matrix. We show how the PARAFAC decomposition can be used to understand the structure of the document space and define paper-paper similarities based on multiple linkages. Examples are presented where the decomposed tensor data is used to find papers similar to a body of work (e.g., related by topic or similar to a particular author's papers), find related authors using linkages other than explicit co-authorship or citations, distinguish between papers written by different authors with the same name, and predict the journal in which a paper was published.
Tensor decompositions (e.g., higher-order analogues of matrix decompositions) are powerful tools for data analysis. In particular, the CANDECOMP/PARAFAC (CP) model has proved useful in many applications such chemometrics, signal processing, and web analysis; see for details. The problem of computing the CP decomposition is typically solved using an alternating least squares (ALS) approach. We discuss the use of optimization-based algorithms for CP, including how to efficiently compute the derivatives necessary for the optimization methods. Numerical studies highlight the positive features of our CPOPT algorithms, as compared with ALS and Gauss-Newton approaches.
This report describes the Licensing Support Network (LSN) Assistant--a set of tools for categorizing e-mail messages and documents, and investigating and correcting existing archives of categorized e-mail messages and documents. The two main tools in the LSN Assistant are the LSN Archive Assistant (LSNAA) tool for recategorizing manually labeled e-mail messages and documents and the LSN Realtime Assistant (LSNRA) tool for categorizing new e-mail messages and documents. This report focuses on the LSNAA tool. There are two main components of the LSNAA tool. The first is the Sandia Categorization Framework, which is responsible for providing categorizations for documents in an archive and storing them in an appropriate Categorization Database. The second is the actual user interface, which primarily interacts with the Categorization Database, providing a way for finding and correcting categorizations errors in the database. A procedure for applying the LSNAA tool and an example use case of the LSNAA tool applied to a set of e-mail messages are provided. Performance results of the categorization model designed for this example use case are presented.
The DAKOTA (Design Analysis Kit for Optimization and Terascale Applications) toolkit provides a flexible and extensible interface between simulation codes and iterative analysis methods. DAKOTA contains algorithms for optimization with gradient and nongradient-based methods; uncertainty quantification with sampling, reliability, and stochastic finite element methods; parameter estimation with nonlinear least squares methods; and sensitivity/variance analysis with design of experiments and parameter study methods. These capabilities may be used on their own or as components within advanced strategies such as surrogate-based optimization, mixed integer nonlinear programming, or optimization under uncertainty. By employing object-oriented design to implement abstractions of the key components required for iterative systems analyses, the DAKOTA toolkit provides a flexible and extensible problem-solving environment for design and performance analysis of computational models on high performance computers. This report serves as a developers manual for the DAKOTA software and describes the DAKOTA class hierarchies and their interrelationships. It derives directly from annotation of the actual source code and provides detailed class documentation, including all member functions and attributes.
The DAKOTA (Design Analysis Kit for Optimization and Terascale Applications) toolkit provides a flexible and extensible interface between simulation codes and iterative analysis methods. DAKOTA contains algorithms for optimization with gradient and nongradient-based methods; uncertainty quantification with sampling, reliability, and stochastic finite element methods; parameter estimation with nonlinear least squares methods; and sensitivity/variance analysis with design of experiments and parameter study methods. These capabilities may be used on their own or as components within advanced strategies such as surrogate-based optimization, mixed integer nonlinear programming, or optimization under uncertainty. By employing object-oriented design to implement abstractions of the key components required for iterative systems analyses, the DAKOTA toolkit provides a flexible and extensible problem-solving environment for design and performance analysis of computational models on high performance computers. This report serves as a user's manual for the DAKOTA software and provides capability overviews and procedures for software execution, as well as a variety of example studies.
The DAKOTA (Design Analysis Kit for Optimization and Terascale Applications) toolkit provides a flexible and extensible interface between simulation codes and iterative analysis methods. DAKOTA contains algorithms for optimization with gradient and nongradient-based methods; uncertainty quantification with sampling, reliability, and stochastic finite element methods; parameter estimation with nonlinear least squares methods; and sensitivity/variance analysis with design of experiments and parameter study methods. These capabilities may be used on their own or as components within advanced strategies such as surrogate-based optimization, mixed integer nonlinear programming, or optimization under uncertainty. By employing object-oriented design to implement abstractions of the key components required for iterative systems analyses, the DAKOTA toolkit provides a flexible and extensible problem-solving environment for design and performance analysis of computational models on high performance computers. This report serves as a reference manual for the commands specification for the DAKOTA software, providing input overviews, option descriptions, and example specifications.
We propose two new multilinear operators for expressing the matrix compositions that are needed in the Tucker and PARAFAC (CANDECOMP) decompositions. The first operator, which we call the Tucker operator, is shorthand for performing an n-mode matrix multiplication for every mode of a given tensor and can be employed to concisely express the Tucker decomposition. The second operator, which we call the Kruskal operator, is shorthand for the sum of the outer-products of the columns of N matrices and allows a divorce from a matricized representation and a very concise expression of the PARAFAC decomposition. We explore the properties of the Tucker and Kruskal operators independently of the related decompositions. Additionally, we provide a review of the matrix and tensor operations that are frequently used in the context of tensor decompositions.