The Experts below are selected from a list of 33 Experts worldwide ranked by ideXlab platform

V.k. Garg - One of the best experts on this subject based on the ideXlab platform.

  • On computation of state avoidance control for infinite state systems in Assignment program framework
    IEEE Transactions on Automation Science and Engineering, 2005
    Co-Authors: Ratnesh Kumar, V.k. Garg
    Abstract:

    We study supervisory control of discrete event systems with potentially infinite state-space using state variables for representation and specification. An Assignment program model consisting of state variables and a finite set of Conditional Assignment statements is used for representing a discrete event system, and a predicate over state variables is used for representing a state avoidance control specification. The contribution of this paper is to show how to perform supervisory control computations symbolically. In the case of a Petri net (vector addition system) with the set of forbidden states being a right-closed set, we present a finitely terminating algorithm for maximally permissive supervision. Discrete-event systems are systems with discrete states that evolve in response to discrete events. The state space of such systems can be finite or infinite. The latter case occurs when the state-variables can take infinitely many values such as integers. Certain states in a given system may be "bad", such as a deadlocking state. Then, controllers must be designed to restrict the system behavior by dynamically disabling events occurring in the system so that system never reaches the bad states. Also, certain events can be uncontrollable and cannot be disabled. State-avoidance control of potentially infinite-state discrete-event systems is studied in this paper. A compact program-like modeling formalism has been adopted. Although the control problem for infinite-state systems is in general not solvable in an automated fashion owing to its undecidability established We develop a symbolic technique, that is iterative in nature and can be automated, for computing a control strategy. If the iteration terminates (there is no guarantee though), a controller is computed. We illustrate through several examples where the iterative computation does terminate. Finally, we show that for a certain class of infinite-state systems that can be modeled as Petri nets, the iterative computation is guaranteed to terminate whenever the state-avoidance set is lower-bounded (every state that "dominates" a bad state is itself bad).

Ratnesh Kumar - One of the best experts on this subject based on the ideXlab platform.

  • On computation of state avoidance control for infinite state systems in Assignment program framework
    IEEE Transactions on Automation Science and Engineering, 2005
    Co-Authors: Ratnesh Kumar, V.k. Garg
    Abstract:

    We study supervisory control of discrete event systems with potentially infinite state-space using state variables for representation and specification. An Assignment program model consisting of state variables and a finite set of Conditional Assignment statements is used for representing a discrete event system, and a predicate over state variables is used for representing a state avoidance control specification. The contribution of this paper is to show how to perform supervisory control computations symbolically. In the case of a Petri net (vector addition system) with the set of forbidden states being a right-closed set, we present a finitely terminating algorithm for maximally permissive supervision. Discrete-event systems are systems with discrete states that evolve in response to discrete events. The state space of such systems can be finite or infinite. The latter case occurs when the state-variables can take infinitely many values such as integers. Certain states in a given system may be "bad", such as a deadlocking state. Then, controllers must be designed to restrict the system behavior by dynamically disabling events occurring in the system so that system never reaches the bad states. Also, certain events can be uncontrollable and cannot be disabled. State-avoidance control of potentially infinite-state discrete-event systems is studied in this paper. A compact program-like modeling formalism has been adopted. Although the control problem for infinite-state systems is in general not solvable in an automated fashion owing to its undecidability established We develop a symbolic technique, that is iterative in nature and can be automated, for computing a control strategy. If the iteration terminates (there is no guarantee though), a controller is computed. We illustrate through several examples where the iterative computation does terminate. Finally, we show that for a certain class of infinite-state systems that can be modeled as Petri nets, the iterative computation is guaranteed to terminate whenever the state-avoidance set is lower-bounded (every state that "dominates" a bad state is itself bad).

Peter Schneider-kamp - One of the best experts on this subject based on the ideXlab platform.

  • Optimizing sorting algorithms by using sorting networks
    Formal Aspects of Computing, 2017
    Co-Authors: Michael Codish, Luís Cruz-filipe, Markus Nebel, Peter Schneider-kamp
    Abstract:

    In this paper, we show how the theory of sorting networks can be applied to synthesize optimized general-purpose sorting libraries. Standard sorting libraries are often based on combinations of the classic Quicksort algorithm, with insertion sort applied as base case for small, fixed, numbers of inputs. Unrolling the code for the base case by ignoring loop conditions eliminates branching, resulting in code equivalent to a sorting network. By replacing it with faster sorting networks, we can improve the performance of these algorithms. We show that by considering the number of comparisons and swaps alone we are not able to predict any real advantage of this approach. However, significant speed-ups are obtained when taking advantage of instruction level parallelism and non-branching Conditional Assignment instructions, both of which are common in modern CPU architectures. Furthermore, a close control of how often registers have to be spilled to memory gives us a complete explanation of the performance of different sorting networks, allowing us to choose an optimal one for each particular architecture. Our experimental results show that using code synthesized from these efficient sorting networks as the base case for Quicksort libraries results in significant real-world speed-ups.

  • Applying Sorting Networks to Synthesize Optimized Sorting Libraries
    Logic-Based Program Synthesis and Transformation, 2015
    Co-Authors: Michael Codish, Luís Cruz-filipe, Markus E. Nebel, Peter Schneider-kamp
    Abstract:

    This paper shows an application of the theory of sorting networks to facilitate the synthesis of optimized general purpose sorting libraries. Standard sorting libraries are often based on combinations of the classic Quicksort algorithm with insertion sort applied as the base case for small fixed numbers of inputs. Unrolling the code for the base case by ignoring loop conditions eliminates branching and results in code which is equivalent to a sorting network. This enables the application of further program transformations based on sorting network optimizations, and eventually the synthesis of code from sorting networks. We show that if considering the number of comparisons and swaps then theory predicts no real advantage of this approach. However, significant speed-ups are obtained when taking advantage of instruction level parallelism and non-branching Conditional Assignment instructions, both of which are common in modern CPU architectures. We provide empirical evidence that using code synthesized from efficient sorting networks as the base case for Quicksort libraries results in significant real-world speed-ups.

Michael Codish - One of the best experts on this subject based on the ideXlab platform.

  • Optimizing sorting algorithms by using sorting networks
    Formal Aspects of Computing, 2017
    Co-Authors: Michael Codish, Luís Cruz-filipe, Markus Nebel, Peter Schneider-kamp
    Abstract:

    In this paper, we show how the theory of sorting networks can be applied to synthesize optimized general-purpose sorting libraries. Standard sorting libraries are often based on combinations of the classic Quicksort algorithm, with insertion sort applied as base case for small, fixed, numbers of inputs. Unrolling the code for the base case by ignoring loop conditions eliminates branching, resulting in code equivalent to a sorting network. By replacing it with faster sorting networks, we can improve the performance of these algorithms. We show that by considering the number of comparisons and swaps alone we are not able to predict any real advantage of this approach. However, significant speed-ups are obtained when taking advantage of instruction level parallelism and non-branching Conditional Assignment instructions, both of which are common in modern CPU architectures. Furthermore, a close control of how often registers have to be spilled to memory gives us a complete explanation of the performance of different sorting networks, allowing us to choose an optimal one for each particular architecture. Our experimental results show that using code synthesized from these efficient sorting networks as the base case for Quicksort libraries results in significant real-world speed-ups.

  • Applying Sorting Networks to Synthesize Optimized Sorting Libraries
    Logic-Based Program Synthesis and Transformation, 2015
    Co-Authors: Michael Codish, Luís Cruz-filipe, Markus E. Nebel, Peter Schneider-kamp
    Abstract:

    This paper shows an application of the theory of sorting networks to facilitate the synthesis of optimized general purpose sorting libraries. Standard sorting libraries are often based on combinations of the classic Quicksort algorithm with insertion sort applied as the base case for small fixed numbers of inputs. Unrolling the code for the base case by ignoring loop conditions eliminates branching and results in code which is equivalent to a sorting network. This enables the application of further program transformations based on sorting network optimizations, and eventually the synthesis of code from sorting networks. We show that if considering the number of comparisons and swaps then theory predicts no real advantage of this approach. However, significant speed-ups are obtained when taking advantage of instruction level parallelism and non-branching Conditional Assignment instructions, both of which are common in modern CPU architectures. We provide empirical evidence that using code synthesized from efficient sorting networks as the base case for Quicksort libraries results in significant real-world speed-ups.

Frank Windmeijer - One of the best experts on this subject based on the ideXlab platform.

  • More reliable inference for the dissimilarity index of segregation
    Econometrics Journal, 2015
    Co-Authors: Rebecca Allen, Simon Burgess, Russell Davidson, Frank Windmeijer
    Abstract:

    The most widely used measure of segregation is the so‐called dissimilarity index. It is now well understood that this measure also reflects randomness in the allocation of individuals to units (i.e. it measures deviations from evenness, not deviations from randomness). This leads to potentially large values of the segregation index when unit sizes and/or minority proportions are small, even if there is no underlying systematic segregation. Our response to this is to produce adjustments to the index, based on an underlying statistical model. We specify the Assignment problem in a very general way, with differences in Conditional Assignment probabilities underlying the resulting segregation. From this, we derive a likelihood ratio test for the presence of any systematic segregation, and bias adjustments to the dissimilarity index. We further develop the asymptotic distribution theory for testing hypotheses concerning the magnitude of the segregation index and show that the use of bootstrap methods can improve the size and power properties of test procedures considerably. We illustrate these methods by comparing dissimilarity indices across school districts in England to measure social segregation.

  • More reliable inference for segregation indices
    2009
    Co-Authors: Rebecca Allen, Simon Burgess, Frank Windmeijer
    Abstract:

    most widely used measure of segregation is the dissimilarity index, D. It is now well understood that this measure also reflects randomness in the a llocation of individuals to units; that is, it meas ures deviations from evenness not deviations from randomness. This leads to potentially large values of the segregation index when unit sizes and/or minority p roportions are small, even if there is no underlyin g systematic segregation. Our response to this is to produce an adjustment to the index, based on an underlying statistical model. We specify the assign ment problem in a very general way, with differences in Conditional Assignment probabilities underlying the resulting segregation. From this we derive a likelihood ratio test for the presence of any systematic segregation and a bootstrap bias adjustment to the dissimilarity index. We further d evelop the asymptotic distribution theory for testi ng hypotheses concerning the magnitude of the segregat ion index and show that use of bootstrap methods can improve the size and power properties of test p rocedures considerably. We illustrate these methods by comparing dissimilarity indices across school di stricts in England to measure social segregation.

  • More Reliable Inference for Segregation Indices
    The Centre for Market and Public Organisation, 2009
    Co-Authors: Rebecca Allen, Simon Burgess, Frank Windmeijer
    Abstract:

    The most widely used measure of segregation is the dissimilarity index, D. It is now well understood that this measure also reflects randomness in the allocation of individuals to units; that is, it measures deviations from evenness not deviations from randomness. This leads to potentially large values of the segregation index when unit sizes and/or minority proportions are small, even if there is no underlying systematic segregation. Our response to this is to produce an adjustment to the index, based on an underlying statistical model. We specify the Assignment problem in a very general way, with differences in Conditional Assignment probabilities underlying the resulting segregation. From this we derive a likelihood ratio test for the presence of any systematic segregation and a bootstrap bias adjustment to the dissimilarity index. We further develop the asymptotic distribution theory for testing hypotheses concerning the magnitude of the segregation index and show that use of bootstrap methods can improve the size and power properties of test procedures considerably. We illustrate these methods by comparing dissimilarity indices across school districts in England to measure social segregation.