Recomputation Time Research Articles

Hierarchical Clustering Approach for Selecting Representative Skylines

Recently, the skyline query has attracted interest in a wide range of applications from recommendation systems to computer networks. The skyline query is useful to obtain the dominant data points from the given dataset. In the low-dimensional dataset, the skyline query may return a small number of skyline points. However, as the dimensionality of the dataset increases, the number of skyline points also increases. In other words, depending on the data distribution and dimensionality, most of the data points may become skyline points. With the emergence of big data applications, where the data distribution and dimensionality are a significant problem, obtaining representative skyline points among resulting skyline points is necessary. There have been several methods that focused on extracting representative skyline points with various success. However, existing methods have a problem of re-computation when the global threshold changes. Moreover, in certain cases, the resulting representative skyline points may not satisfy a user with multiple preferences. Thus, in this paper, we propose a new representative skyline query processing method, called representative skyline cluster (RSC), which solves the problems of the existing methods. Our method utilizes the hierarchical agglomerative clustering method to find the exact representative skyline points, which enable us to reduce the re-computation time significantly. We show the superiority of our proposed method over the existing state-of-the-art methods with various types of experiments.

Accelerating the image processing by the optimization strategy for deep learning algorithm DBN

In recent years, image processing especially for remote sensing technology has developed rapidly. In the field of remote sensing, the efficiency of processing remote sensing images has been a research hotspot in this field. However, the remote sensing data has some problems when processing by a distributed framework, such as Spark, and the key problems to improve execution efficiency are data skew and data reused. Therefore, in this paper, a parallel acceleration strategy based on a typical deep learning algorithm, deep belief network (DBN), is proposed to improve the execution efficiency of the DBN algorithm in Spark. First, the re-partition algorithm based on the tag set is proposed to the relief data skew problem. Second, the cache replacement algorithm on the basis of characteristics is proposed to automatic cache the frequently used resilient distributed dataset (RDD). By caching RDD, the re-computation time of frequently reused RDD is reduced, which lead to the decrease of total computation time of the job. The numerical and analysis verify the effectiveness of the strategy.

Time-reversal checkpointing methods for RTM and FWI

Time-domain seismic simulation can form the basis of reverse time depth migration and full-waveform inversion. These applications need to temporally crosscorrelate a forward simulation state with an adjoint simulation state and therefore need to be able to access each time step of a forward simulation in time-reverse order. This requires saving all forward states for all times (which can require more memory than is typically available on a computer system for many problems of interest), or the ability to checkpoint information and rapidly recompute forward simulation states as needed. Prior work has suggested how to do the latter by optimally choosing which forward simulation time steps to checkpoint, thereby enabling the most efficient reuse of memory buffers and minimizing recomputation. The optimal trade-off between memory usage and recomputation can be further improved under the assumption that the information needed to do temporal crosscorrelation is smaller than the information required to restart a simulation from a given time step. This assumption is true for many geophysical problems of interest. The modification can yield a reduction in the memory requirement and recomputation time. The tested examples applied to isotropic elastic reverse time migration and anisotropic viscoelastic full-waveform inversion.

Evaluation of Software-Implemented Fault-Tolerance (SIFT) Approach in Gracefully Degradable Multi-Computer Systems

This paper presents an analytical method for evaluating the reliability improvement for any size of multi-computer system based on Software-Implemented Fault-Tolerance (SIFT). The method is based on the equivalent failure rate Gamma, the single node failure rate lambda, the number of nodes in the system, N, the repair rate mu, the fault coverage factor c, the reconfiguration rate delta, and the percentage of blocking faults b <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> and b <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> . The impact of these parameters on the reliability improvement has been evaluated for a gracefully degradable multi-computer system using our proposed analytical technique based on Markov chains. To validate our approach, we used the SIFT method which implements error detection at the node level, combined with a fast reconfiguration algorithm for avoiding faulty nodes. It is worth noting that the proposed method is applicable to any multi-computer systems' topology. The evaluation work presented in this paper focuses on the combination of analytical and experimental approaches, and more precisely on Markov chains. The SIFT method has been successfully implemented for a multi-computer system, nCube. The time overhead (reconfiguration & recomputation time) incurred by the injected fault, and the fault coverage factor c, are experimentally evaluated by means of a parallel version of the Software Object-Oriented Fault-Injection Tool (nSOFIT). The implemented SIFT approach can be used for real-time applications, when the time constraints should be met despite failures in the gracefully degradable multi-computer system

Reliable multicast routing in mobile networks: a neural-network approach

Mobile networks that handle multicast communication services, such as video-on-demand, news-on-demand etc., require a kind of reliable and secure point-to-point, point-to-multipoint specific group communications for sophisticated organisation of multicast communications. A reliable multicast tree is an efficient connectivity between the source node and the multicast group members through the dependable hosts. Because the mobile host changes its access point over time, multicast routes must be updated. This poses several challenges to provide an efficient multicast routing. A neural-network-based multicast routing algorithm is proposed for constructing a reliable multicast tree that connects the participants of a multicast group. The problem is tackled by dividing a mobile network into clusters of nodes based on the adjacency relation between the nodes (mobile support stations), by considering a suitable neighbourhood distance. The centre cluster, whose nodes are almost equidistant from the multicast group members, is computed to construct a shortest multicast tree that passes through the centre cluster and reliable routers among all the group members. A Kohonens self-organising-map neural network has been used for clustering. Hopfield neural networks are used to construct a multicast tree with a minimum number of links which passes through the nodes that belong to centre cluster. In addition, the tree is constructed as and when the member(s) moves. This scheme should construct a reliable multicast tree and minimise recomputation time of the multicast tree as and when the multicast route is updated when the mobile hosts change their access point due to mobility. The computational power of the proposed neural-network-based multicast routing algorithm is demonstrated through simulation. The algorithm is also tested for mobility of the participating mobile hosts. The proposed work facilitates a possible multicast routing algorithm for future high-speed mobile networks.

Fully dynamic algorithms for permutation graph coloring

We introduce a dynamic model for maintaining permutation graph coloring. Our motivation comes from the strait type river routing problem in VLSI. This paper presents fully dynamic algorithms for the permutation graph coloring problem. These algorithms are designed to handle Insert and Delete operations and answer some queries. The aim is to provide for running times that are asymptotically more efficient than recomputation (off-line algorithms that run in 0(n logw) time, are known [5,6,10,3]). First, the algorithm A^ that runs in 0(n) uniform running time per Insert/Delete operation is presented. Second, a more sophisticated data structure leads to the algorithm A2 that runs in (9(m logw) uniform running time per Insert I Delete, where m denotes the number of chains in the decomposition. It follows from [7,4] that the running time of A2 when the points from the dynamically changing set are drawn independently from a uniform distribution on the unit square is G(yfn logn) per Insert/Delete in probability. Third, we sketch a composite algorithm A3 that switches between A± and A2 guarantees an amortized running time of (min{n,m logw)) per Insert/Delete. Finally, we outline a number of applications

Algorithms of placing recovery points

In this paper we study the complexity of placing recovery points in computer programs, so that the roll-back time is minimized. The roll-back time corresponds to the recomputation time involved from the previous correct point. Here we give O( n 2) time algorithms for placing recovery points when the underlying program model is either a path tree or a rooted tree. We also show that the problem is NP-complete when the underlying program model is a directed graph.

Backtracking in recursive computations

A mathematical (denotational) semantics is constructed for a formalism of recursive equations with the Alternative operator. This formalism enables the combination of recursion and backtracking. The semantics is defined by applying fixpoint theory to set valued functions. We introduce the notion of strategy to produce subsets of the result. Two implementations are suggested using an auxiliary stack, that trade off recomputation time with space in the auxiliary stack. The concept of a sub-fixpoint is introduced, and the implementations are shown to be incomplete even w.r.t. sub-fixpoint values. One special strategy, the leftmost strategy, which stems from problems such as pattern matching or parsing, is discussed.