Science and Technology on Information System Engineering Laboratory

Nanjing, China

Science and Technology on Information System Engineering Laboratory

Nanjing, China
SEARCH FILTERS
Time filter
Source Type

Zhu X.,Nanjing Research Institute of Electronics Technology | Zhao J.,Nanjing Research Institute of Electronics Technology | Zhao J.,Science and Technology On Information System Engineering Laboratory
Proceedings - 2017 32nd Youth Academic Annual Conference of Chinese Association of Automation, YAC 2017 | Year: 2017

Widespread use of mobile terminal has promoted the rapid movements of information service technology. This paper analyzes the shortcomings of service capabilities in the case of the lack of infrastructure, and presents a service persistence method in mobile environment. Based on distributed service deployment, multiple backup and service session switch and other mechanisms, this method effectively protects the service continuity in entire service process. © 2017 IEEE.


Jin X.,Science and Technology on Information System Engineering Laboratory
Proceedings of the IEEE International Conference on Software Engineering and Service Sciences, ICSESS | Year: 2017

Military operational C2 (command & control) requires dynamically focalized information for support. Internet search engines do not meet such requirement, as they don't care who is making what decision, and needs what information for support. It is found that commanders of same type, executing similar tasks, have similar information requirements. By this idea, a knowledge based method is proposed to enable machine to dynamically feel user requirements and serve focalized information. It is proved feasible to improve commander decision making efficiency. © 2016 IEEE.


Xu X.,Science and Technology on Information System Engineering Laboratory | Lu J.,University of Helsinki | Wang W.,Nanjing University
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2016

Classic data analysis techniques generally assume that variables have single values only. However, the data complexity during the age of big data has gone beyond the classic framework such that variable values probably take the form of a set of stochastic measurements instead. We refer to the above case as the stochastic pattern-based symbolic data where each measurement set is an instance of an underlying stochastic pattern. In such a case, non existing classic data analysis approaches, such as the crystal item or fuzzy region ones, could apply yet. For this reason, we put forward a novel I ncremental Hierarchical Clustering algorithm for stochastic Patternbased Symbolic Data (IHCPSD). IHCPSD is robust to overlapping and missing measurements and well adapted for incremental learning. Experiments on synthetic and application on real-life emitter parameter data have validated its effectiveness. © Springer International Publishing Switzerland 2016.


Xu X.,Science and Technology on Information System Engineering Laboratory | Wang W.,Nanjing University | Lu J.,Renmin University of China
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2015

Emitter signal parameter analysis has been widely recognized as one crucial task for communication, electronic reconnaissance and radar intelligence analysis. However, the parameter measurements are characteristic of uncertainty in the form of intervals. In addition, the measurements are typically accumulated continuously. Existing machine learning methods for interval-valued data are unfit in such a case as they generally assume a uniform distribution and are restricted to static data analysis. To address the above problems, we bring forward an incremental class discriminant analysis method on interval-valued emitter signal parameters. Experimental results have validated its effectiveness. © Springer International Publishing Switzerland 2015.


Rao J.,Science and Technology on Information System Engineering Laboratory | Xu X.,Science and Technology on Information System Engineering Laboratory | Wang Q.,Nanjing Electronic Engineering Research Institute
Proceedings of 2015 IEEE International Conference on Communication Software and Networks, ICCSN 2015 | Year: 2015

In order to analyze the impedance characteristics intuitively, the array antenna is equivalent to a port network. Based on this model, the relation between mutual coupling and mutual impedance is obtained, and a way to modify pattern considering mutual coupling is introduced. In addition, multi-population genetic algorithm is applied to compensate pattern considering mutual coupling, so that the mutual coupling effect of antenna array can be eliminated. Finally, all these theoretic research results are proved by simulation. © 2015 IEEE.


Xu X.,Science and Technology on Information System Engineering Laboratory | Wang W.,Nanjing University
International Journal of Pattern Recognition and Artificial Intelligence | Year: 2012

The incremental classifier is superior in saving significant computational cost by incremental learning on continuously increasing training data. However, existing classification algorithms are problematic when applied for incremental learning for multi-class classification. First, some algorithms, such as neural network and SVM, are not inexpensive for incremental learning due to their complex architectures. When applied for multi-class classification, the computational cost would rise dramatically when the class number increases. Second, existing incremental classification algorithms are usually based on a heuristic scheme and sensitive to the training data input order. In addition, in case the test instance is an outlier and belongs to none of the existing classes, few classification algorithms is able to detect it. Finally, the feature selection and weighing schemes being utilized are generally risky for a "siren pitfall" for multi-class classification tasks. To address the above problems, we bring forward an incremental gray relational analysis algorithm (IGRA). Experimental results showed that, when applied for incremental multi-class classification, IGRA is stable in output, robust to training data input order, superior in computational efficiency, and also capable of detecting outliers and alleviating the "siren pitfall". © 2012 World Scientific Publishing Company.


Xu X.,Science and Technology on Information System Engineering Laboratory | Zhang G.,Science and Technology on Information System Engineering Laboratory | Wu W.,Science and Technology on Information System Engineering Laboratory
Lecture Notes in Electrical Engineering | Year: 2015

With the rapid development of data collection and storage technologies, the volume of data is getting so enormous for collection and analysis in a reasonable amount of time. Only a small fraction of the original data could be contained in the databases or data warehouses. Traditional clustering approaches are recognized as an indispensable solution to extract useful knowledge from data. However, existing conventional clustering methods all lack of robustness and computation efficiency when applied on massive data. In this work, we have made several efforts to better address the above problems with novel techniques of automatic window initialization, distribution density threshold, and window traversal based on distribution density. © Springer International Publishing Switzerland 2015.


Xu X.,Science and Technology on Information System Engineering Laboratory | Wang H.,Science and Technology on Information System Engineering Laboratory
Proceedings of SPIE - The International Society for Optical Engineering | Year: 2015

Frequency estimation via signal sorting is widely recognized as one of the most practical technologies in signal processing. However, the estimated frequencies via signal sorting may be inaccurate and biased due to signal fluctuation under different emitter working modes, problems of transmitter circuit, environmental noises or certain unknown interference sources. Therefore, it has become an important issue to further analyze and refine signal frequencies after signal sorting. To address the above problem, we have brought forward an iterative frequency refinement method based on maximum likelihood. Iteratively, the initial estimated signal frequency values are refined. Experimental results indicate that the refined signal frequencies are more informative than the initial ones. As another advantage of our method, noises and interference sources could be filtered out simultaneously. The efficiency and flexibility enables our method to apply in a wide application area, i.e., communication, electronic reconnaissance and radar intelligence analysis. © 2015 SPIE.


Xu X.,Science and Technology on Information System Engineering Laboratory
International Journal of Machine Learning and Cybernetics | Year: 2013

Even though extensive work has been done on clustering gene expression data, none existing algorithms evaluates gene expression coherence simultaneously by both regulation direction and relative proportion. As an example, density-based algorithms group genes with similar expression levels together and may separate genes whose expression levels have a large difference in value but vary in a fixed proportion relative to one another. In order to simultaneously measure profile coherence in regulation proportion as well as regulation direction, we propose a novel tangent transformation method. Experimental results indicate that our tangent transformation method has enhanced the gene expression clustering results significantly. Our tangent transformation method can be flexibly applied for either global clustering or biclustering, in either unsupervised or supervised scenario. © 2012 Springer-Verlag.


Xu X.,Science and Technology on Information System Engineering Laboratory
2016 IEEE 32nd International Conference on Data Engineering Workshops, ICDEW 2016 | Year: 2016

It has been recognized that the variable values of symbolic data may take the form of either an interval or a set of stochastic measurements of some underlying patterns. However, most existing work in symbolic data exploration are still concentrated on interval values only. Although some work in stochastic pattern-based symbolic data has been explored, it is not adaptable enough for various types of symbolic data yet. As a result, we bring forward a novel hierarchical clustering framework for complex symbolic data exploration. Our framework is able to be applied on either qualitative multi-valued or interval-valued or stochastic pattern-based symbolic data. Extensive experiments on a series of synthetic data sets have validated its efficiency and effectiveness. © 2016 IEEE.

Loading Science and Technology on Information System Engineering Laboratory collaborators
Loading Science and Technology on Information System Engineering Laboratory collaborators