Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education

Hefei, China

Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education

Hefei, China
SEARCH FILTERS
Time filter
Source Type

Jia Z.-H.,Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education | Jia Z.-H.,Anhui Science and Technology University | Zhang Y.-L.,Anhui Science and Technology University | Leung J.Y.-T.,Hefei University of Technology | And 2 more authors.
Applied Soft Computing Journal | Year: 2017

We investigate the problem of minimizing the makespan and the total electric power cost simultaneously on a set of parallel identical batch-processing machines, where the jobs with non-identical sizes dynamically arrive. To address the bi-criteria problem, a Pareto-based ant colony optimization (PACO) algorithm is proposed. Depending on whether the current batch being delayed after the job is added into, two candidate lists are constructed to narrow the search space. Moreover, heuristic information is designed for each candidate list to guide the search. In addition, the objective-oriented local optimization methods are applied to improve the solution quality. Finally, the proposed algorithm is compared with existing multi-objective algorithms through extensive simulation experiments. The experimental results indicate that the proposed algorithm outperforms all of the compared algorithms, especially for large-scale problems. © 2017 Elsevier B.V.


Chen J.,Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education | Chen J.,Anhui Science and Technology University | Fang L.,Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education | Fang L.,Anhui Science and Technology University | And 12 more authors.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2017

In network, nodes are joined together in tightly knit groups. Local group information is used to search the natural community. It can be crucial to help us to understand the functional properties of the networks and detect the true community structure. In this paper, we propose an algorithm called Three-way Decision Community Detection Algorithm based on Local Group Information(LGI-TWD) to detect community structure by using local group information. Firstly, we define sub-communities of each node v. Node v and v’s neighbors which are reachable to each other construct one sub-communities of node v. Then, each sub-communities is regarded as a granular, and then hierarchical structure is constructed based on granulation coefficient. Finally, a further classification for boundary region’s nodes can be done according to belonging degree. Compared with other community detection algorithms (N-TWD, CACDA, GN, NFA, LPA), the experimental results on six real world social networks show that LGI-TWD gets higher modularity value Q and more accurate communities. © Springer International Publishing AG 2017.


Chen J.,Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education | Chen J.,Anhui University | Chen J.,Anhui Science and Technology University | Zhao S.,Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education | And 5 more authors.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2015

Three-way decision theory divides all samples into three regions: positive region, negative region and boundary region. A lack of detailed information may make a definite decision impossible for samples in boundary region. These samples may be further handled by using new information. In this paper, we propose a method Multi-View Decision Model based on constructive three-way decision theory. Multi-view Decision Model mines the global information of all samples for decision. All samples firstly are decided by MinCA, which builds the min covers for each class. Then samples in boundary region are classified using Multi-view information. Experiments have shown that in most cases, Multi-View Decision Model is beneficial for reducing boundary region and promoting classification precision. © Springer International Publishing Switzerland 2015.


Zhang Y.,Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education | Zhang Y.,Anhui Science and Technology University | Zou H.,Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education | Zou H.,Anhui Science and Technology University | And 6 more authors.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2014

The concept of three-way decisions was proposed by Yao and most of three-way decisions models use two thresholds α and β to partition the universe into three parts: the positive region POS(X), the negative region NEG(X) and the boundary region BND(X). But acquirement of a pair of thresholds is a challenge. In this paper, we propose a new method to get the thresholds of three-way decisions model. We introduce loss function into three-way decisions model based on constructive covering algorithm (CCA). A loss function is interpreted as the costs of making classification decisions. More specifically, for reducing losses of classification, we change the radius of the cover according to the loss function, then we compute the thresholds based on the modification. This paper propose an effective method to compute thresholds α and β, which is according to cost-sensitive three-way decisions model based on CCA. © 2014 Springer International Publishing.


Xie Y.,Anhui University | Xie Y.,Anhui Science and Technology University | Xie Y.,Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education | Xu R.,Anhui Science and Technology University | And 3 more authors.
Journal of Computers (Finland) | Year: 2013

The paper presents a vision-based natural bare-hand gesture recognition for remote operating LTSS, which offers simple, high resolution and robust against occlusion. The system is a kind of embedded systems that combines the hardware and software design. It allows users to interact with the information by pointing large screen directly with their fingertips approximately parallel to ground. The key technology of the system is fingertip recognition, tracking, computer vision algorithms and operation judgment. Experiments show that our approach extremely suit to be embedded into real-time human-computer interaction systems. © 2013 ACADEMY PUBLISHER.

Loading Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education collaborators
Loading Key Laboratory of Intelligent Computing and Signal Processing of Ministry of Education collaborators