In the context of test data generation, symbolic execution gets more attention as computing power increases continuously. Experiments show that test generation tools based on symbolic execution can get high coverage a...
详细信息
The hardware design of Godson-3A processor adopts the scalable distributed multi-core structure which is based on a 2D mesh. It can make use of multi-chip interconnection to construct a unified topology structure for ...
详细信息
software migration is an effective way to achieve software reuse and reduce the costs of software maintenance. Wrapping approach is a valuable option for software migration. After analyzing the framework of software i...
详细信息
Compared with single-label classification, multi-label classification is more general in practice, since it allows one instance to have more than one label simultaneously. Bayes' Theorem has been successfully appl...
详细信息
Multi-objective optimization (MOP) a fast growing area of research. Bioinformatics data sets come mostly from DNA microarray experiments. The analysis of microarray data sets can provide valuable information on the bi...
详细信息
Formal concept analysis and rough set analysis are two complementary approaches for analyzing data. This paper studies approaches to constructing fuzzy concept lattices based on generalized fuzzy rough approximation o...
详细信息
Formal concept analysis and rough set analysis are two complementary approaches for analyzing data. This paper studies approaches to constructing fuzzy concept lattices based on generalized fuzzy rough approximation operators. For a residual implicator θ satisfying θ (a, b) = θ(1-b, 1-a) and its dual σ, a pair of (θ, σ)-fuzzy rough approximation operators is defined. We then propose three kinds of fuzzy operators, and examine some of their basic properties. Thus, three complete fuzzy concept lattices can be produced, for which the properties are analogous to those of the classical concept lattices.
Based on the analysis of fault location methods, this paper presents an intelligent method on the basis of classification in large-scale measurement system. Three independent functional units, which are data acquisiti...
详细信息
Based on the analysis of fault location methods, this paper presents an intelligent method on the basis of classification in large-scale measurement system. Three independent functional units, which are data acquisition unit (DAU), data transmission unit (DTU) and terminal computer (TC), are defined in distributed environment. And then, a fault localization method corresponding to all levels are proposed in detail. Finally we conduct experiments on our experimental platform, and results show that our approach can accurately and quickly achieve the mission of fault location in large-scale distributed measurement system.
Neighborhood-based Collaborative Filtering (CF) is one of the most popular techniques applied in recommendation systems. It can be further divided into user-based and item-based CF algorithms. Traditionally, when the ...
详细信息
Neighborhood-based Collaborative Filtering (CF) is one of the most popular techniques applied in recommendation systems. It can be further divided into user-based and item-based CF algorithms. Traditionally, when the two algorithms are evaluated, it is often assumed that the dataset composed of user ratings are free of noise, however, ratings are prone to noise, thus, evaluation results obtained without considering the data noise may be deviated or even be distorted. For a more objective evaluation of the two algorithms, in this paper, we re-evaluate them on latent preference, a more accurate representation of user preference (than rating) based on psychometric models. Experimental results show some differences from those reported in the literature, but the results conform more to common sense. The implications of the findings and the reason for the traditional counter-intuitive results are presented at last.
Parallel file systems suffer performance degradation of small files generated by many workloads. We provide the systematic analysis of the small files problem from aspects of both parallel file system architecture and...
详细信息
Parallel file systems suffer performance degradation of small files generated by many workloads. We provide the systematic analysis of the small files problem from aspects of both parallel file system architecture and workload characteristics. Based on the analysis, we discuss the deficiencies of current optimizations and propose several open questions and possible research directions to address the problem.
暂无评论