This paper examines task partition problem in a Multiagent based Distributed Open Computing Environment Model(MDOCEM). We first present a formal method to describe the task partition problem, then give a heuristic al...
详细信息
This paper examines task partition problem in a Multiagent based Distributed Open Computing Environment Model(MDOCEM). We first present a formal method to describe the task partition problem, then give a heuristic algorithm to solve the task partition problem that gives an approximate optimum solution.
Feature selection is an integral step of the data mining process to find an optimal subset of features. After examining the problems with both the filter and the wrapper approach to feature selection, we propose a two...
详细信息
Feature selection is an integral step of the data mining process to find an optimal subset of features. After examining the problems with both the filter and the wrapper approach to feature selection, we propose a two-phase (filter and wrapper) feature selection algorithm that can take advantage of both approaches. It begins by running GFSIC (Genetic Feature Selection with Inconsistency Criterion), a filter approach, to remove irrelevant features, then it runs SBFCV (Sensitivity-Based Feature selection with v-fold Cross-Validation), a wrapper approach, to remove redundant or useless features. Analysis and experimental studies show the effectiveness and scalability of the proposed algorithm. The generalization of the neural network is improved when the algorithm is used to pre-process the training data by eliminating irrelevant and useless features from the neural network's consideration.
The Neocognitron, inspired by the mammalian visual system, is a complex neural network with numerous parameters and weights which should be trained in order to utilise it for pattern recognition. However, it is not ea...
详细信息
The Neocognitron, inspired by the mammalian visual system, is a complex neural network with numerous parameters and weights which should be trained in order to utilise it for pattern recognition. However, it is not easy to optimise these parameters and weights by gradient decent algorithms. We present a staged training approach using evolutionary algorithms. The experiments demonstrate that evolutionary algorithms can successfully train the Neocognitron to perform image recognition on real world problems.
A model of intelligent resource management in a large distributed system is proposed. This model not only can efficiently solve the problem concerning the incompletion of the global states and the monotony of the mana...
详细信息
A model of intelligent resource management in a large distributed system is proposed. This model not only can efficiently solve the problem concerning the incompletion of the global states and the monotony of the management tactics, but also has the self-adapting ability in a large distributed system, so that the system can maintain high performance all the time.
Line segment clipping is a basic operation of the visualization process in computer graphics. So far there exist four computational models for clipping a line segment against a window, (1) the encoding, (2) the parame...
详细信息
Line segment clipping is a basic operation of the visualization process in computer graphics. So far there exist four computational models for clipping a line segment against a window, (1) the encoding, (2) the parametric, (3) the geometric transformation, and (4) the parallel cutting. This paper presents an algorithm that is based on the third method. By making use of symmetric properties of a window and transformation operations, both endpoints of a line segment are transformed, so that the basic cases are reduced into two that can be easily handled, thus the problems in NLN and AS where there are too many sulyprocedure calls and basic cases that are difficult to deal with are tackled. Both analytical and experimental results from random input data show that the algorithm is better than other developed ones, in view of the speed and the number of operations.
A cooperative computing model is presented to support the construction of CSCW systems. The model divides a cooperative system into layers and each layer is formed by component objects. A principal focus of the model ...
详细信息
A cooperative computing model is presented to support the construction of CSCW systems. The model divides a cooperative system into layers and each layer is formed by component objects. A principal focus of the model is the provision of an object linking mechanism which links the relevant objects in the same layer of multiple users to realize multi-level collaboration which is independent of the application. In addition, multi-level collaboration obtains fine-grained collaboration and improves collaboration efficiency. The object-based method used ensures a structured collaboration and facilitates cooperative system development.
In this paper, we consider the problem verifying hybrid systems modelled by linear hybrid automata. We extend the traditional regular expressions with time constraints and use them as a language to describe the behavi...
详细信息
The problem of uniqueness of reconstruction in open logic is dealt with by introducing the concept of total ordered partitions, an ordering structure for modeling belief degrees of knowledge, and redefining the recons...
详细信息
The problem of uniqueness of reconstruction in open logic is dealt with by introducing the concept of total ordered partitions, an ordering structure for modeling belief degrees of knowledge, and redefining the reconstruction operation. Based on the resulting definition, a nontrivial condition for the convergency of cognitive processes is given. It is shown that if new knowledge is not always accepted with an extremely skeptical attitude and the changes of belief degrees follow the criterion of minimal changes, the cognitive process will converge. The results provide an approach to unifying two kinds of theories for knowledge base maintenance: belief revision and open logic.
Fuzzy entropy is the measure of fuzziness in a fuzzy set. This paper proposes a kind of generalized fuzzy entropy which may describe various kinds of fuzzy entropy, and discusses its guidance functions in the design o...
详细信息
Fuzzy entropy is the measure of fuzziness in a fuzzy set. This paper proposes a kind of generalized fuzzy entropy which may describe various kinds of fuzzy entropy, and discusses its guidance functions in the design of fuzzy neural network classifiers.
The expansion postponement problem in Pure Type Systems is an open problem raised by R. Pollack in 1992. In this paper, the author presents a set of necessary and sufficient conditions for this problem and a set of su...
详细信息
The expansion postponement problem in Pure Type Systems is an open problem raised by R. Pollack in 1992. In this paper, the author presents a set of necessary and sufficient conditions for this problem and a set of sufficient conditions for it. The author also gives some properties for pure type systems without the expansion rule.
暂无评论