Search

Visual analysis of human motion from video sequences is one of the most active research topics in the field of computer vision. This research has certain practical value and can be widely applied in some places such as: bank,hotel,garage,government department,large public facility. In this paper we present a novel method for judging irregular behavior based on treading track. Firstly, we use an object detection method in which the background subtraction method and the time difference method are averaged by weights to detect moving body, and then we judge whether someone is suspicious or not on the basis of treading track, it improves the judgment ability of irregular behavior recognition and experiment results have shown that this method gives static performances and good robustness. [ABSTRACT FROM AUTHOR]

EAI issue is constantly a significant aspect of enterprise computing area. Furthermore, recent advances in Web Service and integration sever provide a promising pushing to EAI issues. However, few investigations have focused on a general view of EAI technologies and solutions. to provide a clear perspective about EAI, a technology framework-ABCMP is presented in this paper. ABCMP attempts to describe the structure of diverse EAI technologies, which allows a general comprehension of the EAI issue. Moreover, process topic about ABCMP is also discussed to provide a wider vision of the technology framework. [ABSTRACT FROM AUTHOR]

Quantum-behaved Particle Swarm Optimization (QPSO) is a global convergence guaranteed search method, which introduced quantum theory into original Particle Swarm Optimization (PSO). While Simulated Annealing (SA) is another important stochastic optimization with the ability of probabilistic hill-climbing. In this paper, the mechanism of Simulated Annealing is introduced into the weak selection implicit in our QPSO algorithm, which effectively employs both the ability to jump out of the local minima in Simulated Annealing and the capacity of searching the global optimum in QPSO algorithm. The experimental results show that the proposed hybrid algorithm increases the diversity of the population in the search process and improves its precision in the latter period of the search. [ABSTRACT FROM AUTHOR]

This paper designs a new kind of structured population and evolutionary operators to form a novel algorithm, Organizational Evolutionary Algorithm (OEA), for solving constrained optimization problems. A simple and non problem-dependent technique is incorporated into OEA to handle the constraints. In OEA, a population consists of organizations, and an organization consists of individuals. All evolutionary operators are designed to simulate the interaction among organizations. In experiments, 4 well-studied engineering design problems are used to test the performance of OEA. The results show that OEA obtains good results both in the solution quality and the computational cost. Keywords: Evolutionary algorithms, organization, constrained optimization. [ABSTRACT FROM AUTHOR]

QoS multicast routing in networks is a very important research issues in the areas of networks and distributed systems. Because of its NP-completeness, many heuristics such as Genetic Algorithms (GAs) are employ solve the QoS routing problem. Base on the previously proposed Quantum-behaved Particle Swarm Optimization (QPSO), this paper proposes a QPSO-based QoS multicast routing algorithm. The proposed method converts the QoS multicast routing problem into an integer programming problem and then solve the problem by QPSO. We test QPSO-base routing algorithm on a network model. For performance comparison, we also test Particle Swarm Optimization (PSO) algorithm and GA. The experiment results show the availability and efficiency of QPSO on the problem and its superiority to PSO and GA. [ABSTRACT FROM AUTHOR]

This paper presents a method of parameterization for online freehand drawing objects based on a piecewise cubic Bezier curve approximation. The target is to represent sketches in a compact format within a certain error tolerance with lower computation to be practically adaptable for the online graphics input. A set of user's intended breakpoints in digital ink is firstly produced in terms of pen speed and local curvatures. Each of strokes of a skechy shape is then parameterized by the optimization of piecewise Bezier curve approximation to minimize the fitting error between stroke path and the curve. The experimental results show both effective and efficient for a wide range of drawing graphic objects. [ABSTRACT FROM AUTHOR]

In a complex environment where the messages exchange tensely among the agents, a difficulty task is to decide the best action for new arriving messages during on-line control. The Meta-Level Control model is modified and used to improve the performance of the communication among the agents in this research. During the control process, the decision is made from the experience acquired by the agents with reinforcement learning. The research proposed a Messages Meta Manager (MMM) model for Air Flow Management System (AFMS) with the combination of the Meta-Level Control approach and reinforcement learning algorithms. With the developed system, the cases of initial heuristic (IH), epsilon adaptative (EA) and performance heuristic (PH) were tested. The results from simulation and analyses show the satisfactory to the research purpose. Keywords: Air Traffic Flow Management, Meta-Level Control, Reinforcement learning. [ABSTRACT FROM AUTHOR]

SB-Tree is a data structure proposed to represent time series according to the importance of the data points. Its advantages over traditional time series representation approaches include: representing time series directly in time domain (shape preservation), retrieving time series data according to the importance of the data points and facilitating multi-resolution time series retrieval. Based on these benefits, one may find this representation particularly attractive in financial time series domain and the corresponding data mining tasks, i.e. categorization and clustering. In this paper, an investigation on the size of the SB-Tree is reported. Two SB-Tree optimization approaches are proposed to reduce the size of the SB-Tree while the overall shape of the time series can be preserved. As demonstrated by various experiments, the proposed approach is suitable for different categorization and clustering applications. [ABSTRACT FROM AUTHOR]

Data mining tools generally deal with highly structured and precise data. However, classical methods fail to handle imprecise or uncertain information. This paper proposes a neuro-fuzzy data mining approach which provides a means to deal with the uncertainty of data. This presents a location-based service collaboration framework and uses the neuro-fuzzy algorithm for data mining. It also introduces the user-profile frequency count (UFC) function to determine the relevance of the information to mobile users. The result of using neuro-fuzzy system provides comprehensive and highly accurate rules. [ABSTRACT FROM AUTHOR]

The main objective of this work is to put forward a chance constrained mixed-integer nonlinear fuzzy programming model for refinery short-term crude oil scheduling problem under demands uncertainty of distillation units. The model studied has characteristics of discrete events and continuous events coexistence, multistage, multiproduct, uncertainty and large scale. Firstly, the model is transformed into its equivalent fuzzy mixed-integer linear programming model by using the method of Quesada & Grossmann [5]. Then the fuzzy equivalent model is changed into its crisp MILP model relies on the theory presented by Liu & Iwamura [12] for the first time in this area. Finally, the crisp MILP model is solved in LINGO 8.0 based on time discretization. A case study which has 265 continuous variables, 68 binary variables and 318 constraints is effectively solved with the proposed solution approach. [ABSTRACT FROM AUTHOR]

Siguo game is a fascinating imperfect information game that provides a new testbed for AI. We have written a computer program that plays Siguo game. This paper reveals and discusses the method that is based on optimistic and cautious attitudes to construct evaluation function of our system. We also present the worth and rank of material and analyze several features of evaluation function, which are piece capture, position domain of piece and oriflamme guard. Each feature of evaluation function is evaluated by general and optimistic algorithm, respectively. [ABSTRACT FROM AUTHOR]

This study proposes a novel method for business forecasting based on fuzzy support vector machines regression (FSVMR). By an application on sales forecasting, details of proposed method are presented including data preprocessing, kernel selection, parameters tuning and so on. The experimental result shows the method's validity. [ABSTRACT FROM AUTHOR]

Structure health monitoring aims to detect the nature of structure damage by using a network of sensors, whose sensor signals are highly correlated and mixed with noise, it is difficult to identify direct relationship between sensors and abnormal structure characteristics. In this study, we apply sensor sensitivity analysis on a structure damage identifier, which integrates independent component analysis (ICA) and support vector machine (SVM) together. The approach is evaluated on a benchmark data from University of British Columbia. Experimental results show sensitivity analysis not only helps domain experts understand the mapping from different location and type of sensors to a damage class, but also significantly reduce noise and improve the accuracy of different level damages identification. [ABSTRACT FROM AUTHOR]

Organizational Co-Evolutionary Classification (OCEC) is a novel classification algorithm, based on co-evolutionary computation. Differing from Genetic Algorithm, OCEC can work without encoding datasets because introducing "organization" concept. To deal with mass data in intrusion detection effectively, we develop a new algorithm, Clustering Organizational Co-Evolutionary Classification (COCEC) by introducing the clustering method to OCEC. COCEC divides initial data into many sections, and each section is considered as an organization, thus COCEC allows more data to obtain evolutionary learning, so the rule set worked out by COCEC contains fewer rules. In addition to improvement of the initial state in OCEC, some improvements have also been done in the choice strategy of the operators and the rule matching method The experiment results show that COCEC is more accurate and more effective than OCEC and OCEFC (Organizational Co-Evolutionary Fuzzy Classification) with the KDD CUP 99 database, and it greatly reduces the number of rules and testing time. [ABSTRACT FROM AUTHOR]

Outsourcing information technology (IT) is a major contemporary strategic decision. This paper proposes an intelligent decision support framework for effective IT outsourcing management. The proposed framework uses case-based reasoning as the main intelligent technique and integrates rule-base reasoning and compromise programming techniques in fuzzy environment for a real-time decision-making. While integrating different methodologies, our motaivation is to take the advantage of their strengths and cancel out each other's weaknesses. The framework potentially leads to more accurate, flexible and efficient retrieval of alternatives that are most similar and most useful to the current decision situation. Finally, a real-life case is given to validate the feasibility of the proposed framework. [ABSTRACT FROM AUTHOR]

We propose a method of multiple context fusion based robust face detection scheme, multiple cascade and finally decision using correlation table. It takes advantage of multiple cascade face detector fusion by context. We propose the filtering classifier method for illumination face image. And then we constructed cascade classifier from applied different filtering method. The multiple cascade detectors made from six single context detectors. Six contexts are divided k-means algorithm, and classify illuminant. In this paper, we proposed the classifier fusion method by using correlation between face images. The proposed face detection achieves the capacity of the high level attentive process by taking advantage of the context-awareness using the information from illumination. We achieved very encouraging experimental results having varying illuminant. [ABSTRACT FROM AUTHOR]

Various approaches are presented to solve the growing spam problem. However, most of these approaches are inflexible to adapt to spam dynamically. This paper proposes a novel approach to counter spam based on spam behavior recognition using Decision Tree learned from data maintained during transfer sessions. A classification is set up according to email transfer patterns enabling normal servers to detect malicious connections before mail body delivered, which contributes much to save network bandwidth wasted by spams. An integrated Anti-Spam framework is founded combining the Behavior Classification with a Bayesian classification. Experiments show that the Behavior Classification has high precision rate with acceptable recall rate considering its bandwidth saving feature. The integrated filter has a higher recall rate than either of the sub-modules, and the precision rate remains quite close to the Bayesian Classification. [ABSTRACT FROM AUTHOR]

Biomedical named entity recognition is a critical task for automatically mining knowledge from biomedical literature. In this paper, we introduce Conditional Random Fields model to recognize biomedical named entities from biomedical literature. Rich features including literal, context and semantics are involved in Conditional Random Fields model. Shallow syntactic features are first introduced to Conditional Random Fields model and do boundary detection and semantic labeling at the same time, which effectively improve the model's performance. Experiments show that our method can achieve an F-measure of 71.2% in JNLPBA test data and which is better than most of state-of-the-art system. [ABSTRACT FROM AUTHOR]

This paper presents a learning based model for Chinese co-reference resolution, in which diverse contextual features are explored inspired by related linguistic theory. Our main motivation is to try to boost the co-reference resolution performance only by leveraging multiple shallow syntactic and semantic features, which can escape from tough problems such as deep syntactic and semantic structural analysis. Also, reconstruction of surface features based on contextual semantic similarity is conducted to approximate the syntactic and semantic parallel preferences in resolution linguistic theories. Furthermore, we consider two classifiers in the machine learning framework for the co-reference resolution, and performance comparison and combination between them are conducted and investigated. We experimentally evaluate our approaches on standard ACE (Automatic Content Extraction) corpus with promising results. [ABSTRACT FROM AUTHOR]

This paper proposes a multi-feature constrained method (MFC) to acquire co-referent relations from large-scale Chinese corpora. The MFC has two phases: candidate relations extraction and verification. The extraction phase uses distribution distance, pattern homogeneity and coordination distribution features of co-referent target words to extract candidate relations from Chinese corpora. In the verification phase, we define an ontology for co-referent token words, and build a relation graph for all candidate relations. Both the ontology and the graph are integrated to generate individual, joint and reinforced strategies to verify candidate relations. Comprehensive experiments have shown that the MFC is practical, and can also be extended to acquire other types of relations. [ABSTRACT FROM AUTHOR]

We regard answer extraction of Question Answering (QA) system as a classification problem, classifying answer candidate sentences into positive or negative. To confirm the feasibility of this new approach, we first extract features concerning question sentences and answer words from question answer pairs (QA pair), then we conduct experiments based on these features, using Maximum Entropy Model (MEM) as a Machine Learning (ML) technique. The first experiment conducted on the class-TIME_YEAR achieves 81.24% in precision and 78.48% in recall. The second experiment expanded to two other classes-OBJ_SUBSTANCE and LOC_CONTINENT also shows good performance. [ABSTRACT FROM AUTHOR]

There are many observable factors that could influence and determine the time series. The dynamic equations of their interaction are always nonlinear, sometimes chaotic. This paper applied phase space reconstruction method to map time series into multi-dimension space based on chaos theory. Extracted from multi-dimension phase space by the method of sequential deviation detection, outlier set was used to construct a decision tree in order to identify the kinds of outliers. According to the results of decision tree, a trading strategy was set up and applied to Chinese stock market. The results show that, even in bear market, the strategy dictated by decision tree brought in considerable yield. [ABSTRACT FROM AUTHOR]

This paper presents new methodology towards the automatic development of multilingual Web portal for multilingual knowledge discovery and management. It aims to provide an efficient and effective framework for selecting and organizing knowledge from voluminous linguistically diverse Web contents. To achieve this, a concept-based approach that incorporates text mining and Web content mining using neural network and fuzzy techniques is proposed. First, a concept-based taxonomy of themes, which will act as the hierarchical backbone of the Web portal, is automatically generated. Second, a concept-based multilingual Web crawler is developed to intelligently harvest relevant multilingual documents from the Web. Finally, a concept-based multilingual text categorization technique is proposed to organize multilingual documents by concepts. As such, correlated multilingual Web documents can be gathered/filtered/organised/ based on their semantic content to facilitate high-performance multilingual information access. [ABSTRACT FROM AUTHOR]

This paper presents a novel clustering model for mining patterns from imprecise electric load time series. The model consists of three components. First, it contains a process that deals with representation and preprocessing of imprecise load time series. Second, it adopts a similarity metric that uses interval semantic separation (Interval SS)-based measurement. Third, it applies the similarity metric together with the k-means clustering method to construct clusters. The model gives a unified way to solve imprecise time series clustering problem and it is applied in a real world application, to find similar consumption patterns in the electricity industry. Experimental results have demonstrated the applicability and correctness of the proposed model. [ABSTRACT FROM AUTHOR]

Retrieving pertinent parts of a meeting or a conversation recording can help for automatic summarization or indexing of the document. In this paper, we deal with an original task, almost never presented in the literature, which consists in automatically extracting questions utterances from a recording. In a first step, we have tried to develop and evaluate a question extraction system which uses only acoustic parameters and does not need any textual information from a speech-to-text automatic recognition system (called ASR system for Automatic Speech Recognition in the speech processing domain) output. The parameters used are extracted from the intonation curve of the speech utterance and the classifier is a decision tree. Our first experiments on French meeting recordings lead to approximately 75% classification rate. An experiment in order to find the best set of acoustic parameters for this task is also presented in this paper. Finally, data analysis and experiments on another French dialog database show the need of using other cues like the lexical information from an ASR output, in order to improve question detection performance on spontaneous speech. [ABSTRACT FROM AUTHOR]