Browsing by Subject "Computer systems programming"

Now showing 1 - 7 of 7

Open Access
Adaptive ensemble learning with confidence bounds for personalized diagnosis
(AAAI Press, 2016) Tekin, Cem; Yoon, J.; Van Der Schaar, M.
With the advances in the field of medical informatics, automated clinical decision support systems are becoming the de facto standard in personalized diagnosis. In order to establish high accuracy and confidence in personalized diagnosis, massive amounts of distributed, heterogeneous, correlated and high-dimensional patient data from different sources such as wearable sensors, mobile applications, Electronic Health Record (EHR) databases etc. need to be processed. This requires learning both locally and globally due to privacy constraints and/or distributed nature of the multimodal medical data. In the last decade, a large number of meta-learning techniques have been proposed in which local learners make online predictions based on their locally-collected data instances, and feed these predictions to an ensemble learner, which fuses them and issues a global prediction. However, most of these works do not provide performance guarantees or, when they do, these guarantees are asymptotic. None of these existing works provide confidence estimates about the issued predictions or rate of learning guarantees for the ensemble learner. In this paper, we provide a systematic ensemble learning method called Hedged Bandits, which comes with both long run (asymptotic) and short run (rate of learning) performance guarantees. Moreover, we show that our proposed method outperforms all existing ensemble learning techniques, even in the presence of concept drift.
Open Access
An aspect-oriented tool framework for developing process-sensitive embedded user assistance systems
(2011) Tekinerdoǧan, B.; Bozbey, S.; Mester, Y.; Turançiftci, E.; Alkişlar L.
Process-sensitive embedded user assistance aims to provide the end-user the necessary guidance based on the state of the process that is being followed. Unfortunately, the development of these systems is not trivial and has to meet several challenges. The main difficulties appear to be related to integration of process-sensitive guidance in the application and the crosscutting behavior of help concerns. To address these issues we developed an aspect-oriented tool framework Assistant-Pro that can be used to develop process-sensitive embedded user assistance for multiple applications. The framework provides tools for defining the process model, defining guidance related to process steps, and modularizing and weaving help concerns in the target application for which user guidance needs to be provided. The framework has been developed and validated in the context of Aselsan, a large Turkish defense electronics company. © 2011 Springer-Verlag Berlin Heidelberg.
Open Access
Concern-oriented analysis and refactoring of software architectures using dependency structure matrices
(ACM, 2009-03) Tekinerdoǧan, Bedir; Scholten F.; Hofmann, C.; Akşit, M.
Current scenario-based architecture analysis methods analyze the architecture with respect to scenarios that relate to stakeholder concerns. Albeit the primary motivation is to analyze the impact of stakeholders' concerns, it appears that concerns are not explicitly represented as first class abstractions. The lack of an explicit notion of concern in scenario-based analysis approaches can result in an incomplete analysis because scenarios are too specific and can only partially represent the concerns. We propose the concern-oriented architecture analysis method (COSAAM) that builds on scenario-based approaches but includes an explicit notion of concern in the analysis. COSAAM applies Dependency Structure Matrices (DSMs) to represent and analyze the dependencies among scenarios, concerns and architectural elements. Further, COSAAM extends DSMs by introducing explicit DSM patterns and heuristic rules for analyzing the impact of concerns on the architecture and for supporting the refactoring of the architecture. Copyright 2009 ACM.
Open Access
A data-level parallel linear-quadratic penalty algorithm for multicommodity network flows
(Association for Computing Machinery, 1994) Pinar, M. C.; Zenios, S. A.
We describe the development of a data-level, massively parallel software system for the solution of multicommodity network flow problems. Using a smooth linear-quadratic penalty (LQP) algorithm we transform the multicommodity network flow problem into a sequence of independent min-cost network flow subproblems. The solution of these problems is coordinated via a simple, dense, nonlinear master program to obtain a solution that is feasible within some user-specified tolerance to the original multicommodity network flow problem. Particular emphasis is placed on the mapping of both the subproblem and master problem data to the processing elements of a massively parallel computer, the Connection Machine CM-2. As a result of this design we can solve large and sparse optimization problems on current SIMD massively parallel architectures. Details of the implementation are reported, together with summary computational results with a set of test problems drawn from a Military Airlift Command application.
Open Access
Distributed evaluation of an iterative function for all object pairs on an SIMD hypercube
(Elsevier BV, 1991) Erçal, F.
An efficient distributed algorithm for evaluating an iterative function on all pairwise combinations of C objects on an SIMD hypercube is presented. The algorithm achieves uniform load distribution and minimal, completely local interprocessor communication. © 1991.
Open Access
Managing aspect orderings to support multiple quality concerns
(IEEE, 2009-03) Tekinerdoǧan, B.; Er, E.
When multiple aspects are composed undesired behavior may emerge due to the interference of aspects. Different interference management approaches have been proposed including detection and resolution of the conflicting aspects. It appears that the majority of the existing approaches have basically focused on functional correctness, whereby orderings of aspects are evaluated with respect to assumed contractual specification. Although functional correctness is an important quality concern also other quality concerns such as evolvability, reuse and reliability can demand a specific ordering. As such, the resulting possible set of orderings might need to be further reduced. In this paper we discuss the impact of other quality concerns than functional correctness, on the required orderings of aspects. Based on a domain analysis of existing approaches we provide a feature model and complementary to this a metamodel for defining aspect interference management approaches for multiple quality concerns. Copyright 2009 ACM.
Open Access
Voltage island based heterogeneous NoC design through constraint programming
(Pergamon Press, 2014) Demiriz, A.; Bagherzadeh, N.; Ozturk, O.
This paper discusses heterogeneous Network-on-Chip (NoC) design from a Constraint Programming (CP) perspective and extends the formulation to solving Voltage-Frequency Island (VFI) problem. In general, VFI is a superior design alternative in terms of thermal constraints, power consumption as well as performance considerations. Given a Communication Task Graph (CTG) and subsequent task assignments for cores, cores are allocated to the best possible places on the chip in the first stage to minimize the overall communication cost among cores. We then solve the application scheduling problem to determine the optimum core types from a list of technological alternatives and to minimize the makespan. Moreover, an elegant CP model is proposed to solve VFI problem by mapping and grouping cores at the same time with scheduling the computation tasks as a limited capacity resource allocation model. The paper reports results based on real benchmark datasets from the literature.