Formal techniques for performance analysis: blending SAN and PEPA

In this paper we consider two performance modelling techniques from the perspectives of model construction, generation of an underlying continuous time Markov process, and the potential for reduction in the Markov process. Such careful comparison of modelling techniques allows us to appreciate the strengths and weaknesses of different approaches, and facilitates cross-fertilization between them. In the present case we take a characteristic of one formalism, functional rates in Stochastic Automata Networks, and introduce it to the other formalism, Performance Evaluation Process Algebra. We investigate the benefits of this cross-fertilization, particularly from the perspectives of Markov process generation and reduction.     

Polite Combination of Algebraic Datatypes

Journal of Automated Reasoning - Algebraic datatypes, and among them lists and trees, have attracted a lot of interest in automated reasoning and Satisfiability Modulo Theories (SMT). Since its...     

Adapting application execution in CMPs using helper threads

In parallel to the changes in both the architecture domain–the move toward chip multiprocessors (CMPs)–and the application domain–the move toward increasingly data-intensive workloads–issues such as performance, energy efficiency and CPU availability are becoming increasingly critical. The CPU availability can change dynamically due to several reasons such as thermal overload, increase in transient errors, or operating system scheduling. An important question in this context is how to adapt, in a CMP, the execution of a given application to CPU availability change at runtime. Our paper studies this problem, targeting the energy-delay product (EDP) as the main metric to optimize. We first discuss that, in adapting the application execution to the varying CPU availability, one needs to consider the number of CPUs to use, the number of application threads to accommodate and the voltage/frequency levels to employ (if the CMP has this capability). We then propose to use helper threads to adapt the application execution to CPU availability change in general with the goal of minimizing the EDP. The helper thread runs parallel to the application execution threads and tries to determine the ideal number of CPUs, threads and voltage/frequency levels to employ at any given point in execution. We illustrate this idea using four applications (Fast Fourier Transform, MultiGrid, LU decomposition and Conjugate Gradient) under different execution scenarios. The results collected through our experiments are very promising and indicate that significant EDP reductions are possible using helper threads. For example, we achieved up to 66.3%, 83.3%, 91.2%, and 94.2% savings in EDP when adjusting all the parameters properly in applications FFT, MG, LU, and CG, respectively. We also discuss how our approach can be extended to address multi-programmed workloads.     

Improvement of the kernel minimum squared error model for fast feature extraction

The kernel minimum squared error (KMSE) expresses the feature extractor as a linear combination of all the training samples in the high-dimensional kernel space. To extract a feature from a sample, KMSE should calculate as many kernel functions as the training samples. Thus, the computational efficiency of the KMSE-based feature extraction procedure is inversely proportional to the size of the training sample set. In this paper, we propose an efficient kernel minimum squared error (EKMSE) model for two-class classification. The proposed EKMSE expresses each feature extractor as a linear combination of nodes, which are a small portion of the training samples. To extract a feature from a sample, EKMSE only needs to calculate as many kernel functions as the nodes. As the nodes are commonly much fewer than the training samples, EKMSE is much faster than KMSE in feature extraction. The EKMSE can achieve the same training accuracy as the standard KMSE. Also, EKMSE avoids the overfitting problem. We implement the EKMSE model using two algorithms. Experimental results show the feasibility of the EKMSE model.     

A visual language for modeling and executing traceability queries

Current software and systems engineering tools provide only basic trace features, and as a result users are often compelled to construct non-trivial traceability queries using generic query languages such as SQL. In this paper, we present an alternative approach which defines traceability strategies for a project using UML class diagrams and then constructs trace queries as constraints upon subsets of the model. The visual trace modeling language (VTML) allows users to model a broad range of trace queries while hiding underlying technical details and data structures. The viability and expressiveness of VTML for use in a real project are demonstrated through modeling a broadly representative set of queries for a web-based health-care system. It is then evaluated through an experiment with human users to assess the readability and writability of VTML queries in comparison to generic SQL queries. We found that users read and constructed traceability queries considerably faster using VTML than using SQL. Furthermore, visually constructed traceability queries were substantially more correct compared to the same queries constructed with SQL.     

Effects of hostility on ambulatory blood pressure and mood during daily living in healthy adults.

This study (a) tested the effects of hostile attributes on ambulatory blood pressure (BP), heart rate, and mood monitored repeatedly over 3 days in 100 healthy men and women and (b) determined whether the cardiovascular effects of trait hostility were moderated by mood. Multilevel random-coefficients regression analyses showed that hostile individuals exhibited higher systolic and diastolic BP and rated their current moods as more negative and less positive throughout the monitoring. Individuals low in hostility exhibited high BP only during the few occasions when they experienced negative mood. However, these patterns were true only when participants were classified by Potential for Hostility ratings from the Structured Interview (R. H. Rosenman, 1978), not by the Cynical Hostile Attitudes score derived from the Cook-Medley scale. Results provide convergent and ecological validity of interview rating of hostility and illuminate one possible dynamic mechanism by which overt hostile behaviors might contribute to the rates of increased cardiovascular morbidity and mortality. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Typed Abstract State Machines for data-intensive applications

This paper focuses on extensions of Abstract State Machines (ASMs) for the benefit of a refinement calculus in the area of data-intensive applications, in particular data warehouses and on-line analytical processing. The extensions lead to a typed version of ASMs and a more specific notion of strong data refinement, which incorporates the preservation of information content by means of schema dominance. Each typed ASM can be effectively translated into an equivalent ordinary one. Providing typed ASMs helps to exploit the existing logical formalisms used in data-intensive applications to define a ground model. Furthermore, strong data refinement helps to set up a refinement-based development method that uses provably correct standard refinement rules.     

Progression into and out of binge drinking among high school students.

The current study examined binge drinking among high school students over an academic year. Adolescent drinkers (N?=?621; 58% female) were grouped into 4 trajectories: drinkers (35%), increasers (14%), decreasers (16%), and persistent binge drinkers (35%). Prospective analyses indicated several factors that predicted escalation and de-escalation of binge drinking. Increasers were more likely to regularly use alcohol and cigarettes at a younger age than drinkers. Compared with decreasers, persistent binge drinkers reported regular alcohol and marijuana use at younger ages. Lower levels of perceived student drinking appeared to be a protective factor for onset of binge drinking. The results highlight the need to study precursors to the naturally occurring fluctuations in binge drinking and suggest factors that may accentuate the risk of binge drinking. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Prediction of polymer tube life for solar hot water systems: A model of antioxidant loss

The successful use of polymer components in solar hot water systems requires an understanding and method of predicting material degradation in water. The relevant degradation mechanism is oxidation. Degradation from oxidation can be delayed through the use of antioxidant additives. However, once the antioxidants have been depleted, oxidation of the polymer and subsequent loss of mechanical integrity occur rapidly. In this study, antioxidant loss from polymer tubes is modeled. Dimensional analysis and results of the model show the rate of antioxidant loss is controlled by diffusion through the polymer. The diffusion time scale is dictated by the tube wall thickness. Antioxidant concentration profiles and depletion rates are presented for three representative tube geometries and temperatures of 293 and 333 K. The time to deplete 90% of the antioxidant is on the order of 1000–10,000 h.