The Experts below are selected from a list of 111 Experts worldwide ranked by ideXlab platform
Xin Chen - One of the best experts on this subject based on the ideXlab platform.
-
tolerating temporal correlated failures from Cyclic Dependency in high performance computing systems
International Conference on Parallel and Distributed Systems, 2008Co-Authors: Xin ChenAbstract:Correlated failures have recently gained more attention in the research of failures in large scale systems. Recent studies have pointed out the negative effect of ignoring such failures when designing a fault tolerant scheme for large scale systems. In this paper, we explore the behaviors of temporal correlated failures arising from Cyclic Dependency among task nodes via an abstract model. Using this model, we find that fast failure propagation and slow recovery from failures are two dominant factors which make recovering from such failures much difficult. To efficiently stop failure propagation and shorten the total recovering time, we propose a recovery protocol called GCCTS (group-based coordinated checkpointing and task suspending) against temporal correlated failures.
-
ICPADS - Tolerating Temporal Correlated Failures from Cyclic Dependency in High Performance Computing Systems
2008 14th IEEE International Conference on Parallel and Distributed Systems, 2008Co-Authors: Xin ChenAbstract:Correlated failures have recently gained more attention in the research of failures in large scale systems. Recent studies have pointed out the negative effect of ignoring such failures when designing a fault tolerant scheme for large scale systems. In this paper, we explore the behaviors of temporal correlated failures arising from Cyclic Dependency among task nodes via an abstract model. Using this model, we find that fast failure propagation and slow recovery from failures are two dominant factors which make recovering from such failures much difficult. To efficiently stop failure propagation and shorten the total recovering time, we propose a recovery protocol called GCCTS (group-based coordinated checkpointing and task suspending) against temporal correlated failures.
Yoram Singer - One of the best experts on this subject based on the ideXlab platform.
-
feature rich part of speech tagging with a Cyclic Dependency network
North American Chapter of the Association for Computational Linguistics, 2003Co-Authors: Kristina Toutanova, Dan Klein, Christopher D Manning, Yoram SingerAbstract:We present a new part-of-speech tagger that demonstrates the following ideas: (i) explicit use of both preceding and following tag contexts via a Dependency network representation, (ii) broad use of lexical features, including jointly conditioning on multiple consecutive words, (iii) effective use of priors in conditional loglinear models, and (iv) fine-grained modeling of unknown word features. Using these ideas together, the resulting tagger gives a 97.24% accuracy on the Penn Treebank WSJ, an error reduction of 4.4% on the best previous single automatically learned tagging result.
-
HLT-NAACL - Feature-rich part-of-speech tagging with a Cyclic Dependency network
Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology - NAACL, 2003Co-Authors: Kristina Toutanova, Dan Klein, Christopher D Manning, Yoram SingerAbstract:We present a new part-of-speech tagger that demonstrates the following ideas: (i) explicit use of both preceding and following tag contexts via a Dependency network representation, (ii) broad use of lexical features, including jointly conditioning on multiple consecutive words, (iii) effective use of priors in conditional loglinear models, and (iv) fine-grained modeling of unknown word features. Using these ideas together, the resulting tagger gives a 97.24% accuracy on the Penn Treebank WSJ, an error reduction of 4.4% on the best previous single automatically learned tagging result.
Kristina Toutanova - One of the best experts on this subject based on the ideXlab platform.
-
feature rich part of speech tagging with a Cyclic Dependency network
North American Chapter of the Association for Computational Linguistics, 2003Co-Authors: Kristina Toutanova, Dan Klein, Christopher D Manning, Yoram SingerAbstract:We present a new part-of-speech tagger that demonstrates the following ideas: (i) explicit use of both preceding and following tag contexts via a Dependency network representation, (ii) broad use of lexical features, including jointly conditioning on multiple consecutive words, (iii) effective use of priors in conditional loglinear models, and (iv) fine-grained modeling of unknown word features. Using these ideas together, the resulting tagger gives a 97.24% accuracy on the Penn Treebank WSJ, an error reduction of 4.4% on the best previous single automatically learned tagging result.
-
HLT-NAACL - Feature-rich part-of-speech tagging with a Cyclic Dependency network
Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology - NAACL, 2003Co-Authors: Kristina Toutanova, Dan Klein, Christopher D Manning, Yoram SingerAbstract:We present a new part-of-speech tagger that demonstrates the following ideas: (i) explicit use of both preceding and following tag contexts via a Dependency network representation, (ii) broad use of lexical features, including jointly conditioning on multiple consecutive words, (iii) effective use of priors in conditional loglinear models, and (iv) fine-grained modeling of unknown word features. Using these ideas together, the resulting tagger gives a 97.24% accuracy on the Penn Treebank WSJ, an error reduction of 4.4% on the best previous single automatically learned tagging result.
Christopher Gill - One of the best experts on this subject based on the ideXlab platform.
-
IPDPS - On efficient distributed deadlock avoidance for real-time and embedded systems
Proceedings 20th IEEE International Parallel & Distributed Processing Symposium, 2006Co-Authors: César Sánchez, Henny B. Sipma, Zohar Manna, Venkita Subramonian, Christopher GillAbstract:Thread allocation is an important problem in distributed real-time and embedded (DRE) systems. A thread allocation policy that is too liberal may cause deadlock, while a policy that is too conservative limits potential parallelism, thus wasting resources. However, achieving (globally) optimal thread utilization, while avoiding deadlock, has been proven impractical in distributed systems: it requires too much communication between components. In previous work we showed that efficient local thread allocation protocols are possible if the protocols are parameterized by global static data, in particular by an annotation of the global call graph of all tasks to be performed by the system. We proved that absence of Cyclic dependencies in this annotation guarantees absence of deadlock. In this paper we present an algorithm to compute optimal annotations, that is annotations that maximize parallelism while satisfying the condition of aCyclicity. Moreover, we show that the condition of aCyclicity is in fact tight and exhibits a rather surprising anomaly: if a Cyclic Dependency is present in the annotation of the call graph and a certain minimum number of threads is provided, deadlock is reachable. Thus, in the presence of Cyclic dependencies, increasing the number of threads may introduce the possibility of deadlock in an originally deadlock free system.
Julien Schmaltz - One of the best experts on this subject based on the ideXlab platform.
-
Proof Pearl: A Formal Proof of Dally and Seitz’ Necessary and Sufficient Condition for Deadlock-Free Routing in Interconnection Networks
Journal of Automated Reasoning, 2012Co-Authors: Freek Verbeek, Julien SchmaltzAbstract:Avoiding deadlock is crucial to interconnection networks. In ’87, Dally and Seitz proposed a necessary and sufficient condition for deadlock-free routing. This condition states that a routing function is deadlock-free if and only if its channel Dependency graph is aCyclic. We formally define and prove a slightly different condition from which the original condition of Dally and Seitz can be derived. Dally and Seitz prove that a deadlock situation induces Cyclic dependencies by reductio ad absurdum . In contrast we introduce the notion of a waiting graph from which we explicitly construct a Cyclic Dependency from a deadlock situation. Moreover, our proof is structured in such a way that it only depends on a small set of proof obligations associated to arbitrary routing functions and switching policies. Discharging these proof obligations is sufficient to instantiate our condition for deadlock-free routing on particular networks. Our condition and its proof have been formalized using the ACL2 theorem proving system.
-
ITP - A formal proof of a necessary and sufficient condition for deadlock-free adaptive networks
Interactive Theorem Proving, 2010Co-Authors: Freek Verbeek, Julien SchmaltzAbstract:Deadlocks occur in interconnection networks as messages compete for free channels or empty buffers. Deadlocks are often associated with a circular wait between processes and resources. In the context of networks, Duato proved that for adaptive routing networks a Cyclic Dependency is not sufficient to create a deadlock. He proposed deadlock-free routing techniques allowing Cyclic dependencies between channels or buffers. His work was a breakthrough. It was also counterintuitive and only a complex mathematical proof could convince his peers about the soundness of his theory. We define a necessary and sufficient condition that captures Duato's intuition but that is more intuitive and leads to a simpler proof. However, our condition is logically equivalent to Duato's one. We used the ACL2 theorem proving system to formalize our condition and its proof. In particular, we used two features of ACL2, namely the encapsulation principle and quantifiers, to perform an elegant formalization based on second order functions.
