Time filter

Source Type

Gao J.S.,Xiamen University | Zhou C.L.,Xiamen University | Zhou C.L.,Fujian Key Laboratory of the Brain like Intelligent Systems
MATEC Web of Conferences | Year: 2016

The process of game is actually the process of game behavior which is often accompanied by agent's epistemic process. And Knowledge is a key factor of epistemology. In the past, most of the time game theory confuses information with knowledge. This paper, based on game behavior, analyzes agent's knowledge differences from information. It also analyzes how agent reasons by hidden knowledge behind game behavior, as well as what kind of knowledge is involved in the process of reasoning, and what is the origin of the knowledge. This paper divides the agent's knowledge into two parts: common knowledge and private knowledge, which analyze the specific elements that are corresponding to them in games. In the end, it constructs a knowledge model of game agent, furthermore, by this model, the essence of knowledge held by agents in their game would be presented. © 2016 The Authors, published by EDP Sciences.


Huang Z.,Xiamen University | Huang Z.,Huaqiao University | Huang Z.,Fujian Key Laboratory of the Brain like Intelligent Systems | Chen Y.,Xiamen University | Shi X.,Xiamen University
Journal of Convergence Information Technology | Year: 2012

Semantic role labeling (SRL) is a fundamental task in natural language processing to find a sentence-level representation. In this paper, a semantic role labeling algorithm based on synergetic neural network is proposed which can easily express useful global constraints. However, its computational complexity is relatively high that restricts its applications in real-time situation. We propose a parallel algorithm and implement it in distributed computing environments based on a local area network. Experimental results show that the parallel algorithm can reduce the run time and has a high speed up.


Lin F.,Fujian Key Laboratory of the Brain Like Intelligent Systems | Lin F.,Xiamen University | Zeng W.,Xiamen University | Xiahou J.,Xiamen University | Jiang Y.,Xiamen University
Journal of Software | Year: 2011

BP artificial neural network is a non-feedback network. This paper utilizes the initial weights of neural network to choose controller performance. Simultaneously according to the characteristics that process of central air-conditioning energy saving control is the system of multi-parameter and nonlinear time-varying complexity, we analysis and study its algorithm and system architecture. The experimental results demonstrate that new control system gets better results and energy saving. © 2011 ACADEMY PUBLISHER.


Huang Z.,Huaqiao University | Huang Z.,Xiamen University | Shi X.,Huaqiao University | Shi X.,Fujian Key Laboratory of the Brain like Intelligent Systems
Proceedings - 2010 IEEE International Conference on Intelligent Computing and Intelligent Systems, ICIS 2010 | Year: 2010

AdaBoost is an excellent machine-learning algorithm, which produces a strong classifier by selecting the discriminating features and combining them linearly. But the computational complexity of AdaBoost algorithms is very high that restricts its studies and applications. In this paper we utilize parallel processing techniques to implement the parallel algorithm based on AdaBoost. The experiment results show the parallel algorithm can reduce the run time and has a high speedup. ©2010 IEEE.


Huang Z.,Huaqiao University | Huang Z.,Xiamen University | Chen Y.,Xiamen University | Chen Y.,Fujian Key Laboratory of the Brain like Intelligent Systems
International Journal of Control and Automation | Year: 2013

The artificial fish swarm algorithm (AFSA) is a heuristic global optimization technique based on population which is easy to understand, good robustness, and not insensitive to initial values. The behavior of fishes has a great impact on the performance of the algorithm, such as global search and convergence speed. At present, there has no general research theory to select behaviors of fishes. In order to deal with this problem, we proposed an improved artificial fish swarm algorithm based on hybrid behavior selection. There are two mainly works in this paper. Firstly, we propose an improved algorithm based on swallowed behavior, which can greatly speed up the convergence. Secondly, in order to deal with the problems of easy to fall into local optimum value, we added breeding behavior to improve global optimization capability. The experiments on high dimensional function optimization showed that the improved algorithm has more powerful global exploration ability and faster convergence speed. Copyright © 2013 SERSC.


Huang Z.,Huaqiao University | Huang Z.,Xiamen University | Chen Y.,Xiamen University | Chen Y.,Fujian Key Laboratory of the Brain Like Intelligent Systems
Computational and Mathematical Methods in Medicine | Year: 2014

Exon recognition is a fundamental task in bioinformatics to identify the exons of DNA sequence. Currently, exon recognition algorithms based on digital signal processing techniques have been widely used. Unfortunately, these methods require many calculations, resulting in low recognition efficiency. In order to overcome this limitation, a two-stage exon recognition model is proposed and implemented in this paper. There are three main works. Firstly, we use synergetic neural network to rapidly determine initial exon intervals. Secondly, adaptive sliding window is used to accurately discriminate the final exon intervals. Finally, parameter optimization based on artificial fish swarm algorithm is used to determine different species thresholds and corresponding adjustment parameters of adaptive windows. Experimental results show that the proposed model has better performance for exon recognition and provides a practical solution and a promising future for other recognition tasks. © 2014 Zhehuang Huang and Yidong Chen.


Song Z.,Xiamen University | Song Z.,Fujian Key Laboratory of the Brain like Intelligent Systems | Zhou C.,Xiamen University | Lin K.,Xiamen University
Advances in Information Sciences and Service Sciences | Year: 2011

We introduce a novel and efficient method for highly realistic human skeleton model, which is based on IK (inverse kinematics) and dynamics biomechanical theories. To be built a physically realistic skeleton model as possible as a real human skeleton, we propose the use of dynamic biomechanical to constraints on multi-joint movements and represent joint rotation with the unit quaternion. Finally, we conduct some simulation experiments of human walking on non-flat ground using our skeleton model, and analyze the experimental results.


Gao J.,Xiamen University | Zhou C.,Xiamen University | Zhou C.,Fujian Key Laboratory of the Brain like Intelligent Systems
Proceedings of the 28th Chinese Control and Decision Conference, CCDC 2016 | Year: 2016

The development of Intelligent Decision Support System is a key application for artificial intelligence technology. And human-computer interaction is the most common action in Intelligent Decision Support System. In the process of human-computer interaction, the input of decision problem and the output of the solution will be converted into the knowledge that can be extracted by machine. Human-computer interaction is a dynamic process, which involves knowledge extraction and sharing between machines and people. In this paper, we establish a Knowledge Extraction System (KES) based on the analysis of the usual knowledge types that are involved in the interaction process. And then we prove its soundness and completeness, explore some properties about reasoning in human-computer interaction. In this system, we can find the interpretation as to how the machine understands human's questions and reasons intelligently. It also reflects how the cognitive agent extracts each other's private knowledge into common knowledge that later becomes its own knowledge during the interactive process. © 2016 IEEE.


Huang Z.,Huaqiao University | Huang Z.,Fujian Key Laboratory of the Brain like Intelligent Systems
Sensors and Transducers | Year: 2013

Differential evolution (DE) algorithm is a good optimization technique based on population which has been successfully applied in many research and application areas. Log-linear model is a statistical model which can easily blend multiple features, a variety of knowledge sources can be added to the model in the form of feature functions. Traditional differential evolution algorithm is easy to fall into local optimum value and the convergence rate is slow. To solve these problems, an improved differential evolution algorithm based on loglinear model is proposed and implemented in this paper. There are two mainly works in this paper. Firstly, we introduce log-linear model to differential evolution algorithm which can enhance decision making ability. Secondly, some operations are presented to improve global optimization capability. Experiments showed that the improved algorithm has more powerful global exploration ability and faster convergence speed. © 2013 by IFSA.


Huang Z.,Huaqiao University | Huang Z.,Fujian Key Laboratory of the Brain like Intelligent Systems
International Journal of Emerging Technologies in Learning | Year: 2013

In recent years, with the sustainable and fast development of chinese economy, there has been increasing interest in overseas chinese language. As an important way of overseas chinese education, online education has been drawing more and more attention due to its simplicity and convenience. But at present, the existing chinese education resource service model is simply assigned the resources to users which can not effectively meet the practical needs of users. How to provide a personalized service is the key problem to be solved. In this paper, we proposed a sharing model of chinese education resource based on cloud computing. There are three mainly works in this paper. Firstly, user vector space is constructed based on user personal information. Secondly, synergetic neural network is presented to user group recognition. Finally, sharing model based cloud computing is presented and implemented. The proposed model in this paper provide a good practicability and a promising future for overseas chinese education.

Loading Fujian Key Laboratory of the Brain like Intelligent Systems collaborators
Loading Fujian Key Laboratory of the Brain like Intelligent Systems collaborators