Objective Programming Method for an Intuitionistic Trapezoidal Fuzzy Model (MOPM-ITFM) is proposed to improve the accuracy of estimation by optimizing the parameters. The advantage of the proposal is more realistic to...
详细信息
ISBN:
(纸本)9789813146969
Objective Programming Method for an Intuitionistic Trapezoidal Fuzzy Model (MOPM-ITFM) is proposed to improve the accuracy of estimation by optimizing the parameters. The advantage of the proposal is more realistic to consider the degrees of both the acceptance and the rejection of intuitionistic trapezoidal fuzzy number (ITFN) than conventional fuzzy model only considering the former. The proposal is realized by tuning the membership function shapes, the adaptive factors of matching degree and the rule weights. The constraint condition of the proposal is considered in the definition of ITFN. Input fuzzy partitions of an Intuitionistic Trapezoidal Fuzzy Model (ITFM) are modified through the proposal. MOPM-ITFM has been used for a medical diagnosis. Compared with Single Objective Programming Method (SOPM), the results show that the mean square error (MSE) between the output given by domain experts and the output of (MOPM-ITFM) improves 47% than that of SOPM. In the further work, the interpretability of input fuzzy partitions is also needed to be improved in ITFM.
The concept of Key Performance Indicators (KPI) facing the personalized garment design and production process is proposed in this paper. The way of establishing the Key Performance Indicators (KPI) for measuring the p...
详细信息
ISBN:
(纸本)9789813146969
The concept of Key Performance Indicators (KPI) facing the personalized garment design and production process is proposed in this paper. The way of establishing the Key Performance Indicators (KPI) for measuring the performance of the personalized garment design and production is discussed. The application of these KPIs is realized given a reasonable weight, which is obtained by the FANP methods.
This paper describes an approach for the problem of buttock shape discrimination using support vector machines (SVM). Characteristics of female buttock shapes were analyzed through elliptic Fourier analysis and 13 fea...
详细信息
ISBN:
(纸本)9789813146969
This paper describes an approach for the problem of buttock shape discrimination using support vector machines (SVM). Characteristics of female buttock shapes were analyzed through elliptic Fourier analysis and 13 feature vectors were calculated to represent buttock shapes. Then three kinds of 317 Chinese female buttock shapes were selected as samples. The training set included 70% of the samples, while the rest 30% of samples were in prediction set. With the method of K-fold cross validation, the best values of parameters used in SVM were calculated. Finally, it achieved a high accuracy of more than 96% in the discriminating test. The results can be used as an effective way to help clothing companies to subdivide goal customer groups based on the features of their buttock shapes.
In this paper, a performance measure framework is presented for the quality evaluation of government websites. A review of recent developments in website evaluation is provided to facilitate the assessment and compari...
详细信息
ISBN:
(纸本)9789813146969
In this paper, a performance measure framework is presented for the quality evaluation of government websites. A review of recent developments in website evaluation is provided to facilitate the assessment and comparison of government websites with the help of multi criteria decision making (MCDM) techniques. Government website quality evaluation is typically a hard MCDM problem. On the other hand, MCDM in intuitionistic fuzzy (IF) environment is an effective and commonly applied tool for evaluating problems and comparing results in the presence of a number of usually conflicting decision criteria, for which decision makers have vague and imprecise opinions. In line with several characteristics of website quality, this paper approaches the problem of comparing different websites with an integrated IF AHP and IF TOPSIS MCDM method, and then comes up with a practical and selective approach to solve it. To avoid bias, a group decision making (GDM) approach is applied, which is able to address the difficulty of uncertainty and partiality in the decision making process. A case study of the proposed approach is presented to compare different government websites in Turkey.
In this paper we study problem of fuzzy scheduling with the given number of activities to minimize the project duration, in which the activities are scheduled subject to generalized precedence relations, require some ...
详细信息
ISBN:
(纸本)9789813146969
In this paper we study problem of fuzzy scheduling with the given number of activities to minimize the project duration, in which the activities are scheduled subject to generalized precedence relations, require some units of multiple renewable or non-renewable constrained resources with the limited availability, and can be performed in one of several different ways: multiple activity modes or alternatives. Fuzzy branch and bound solution methodology was presented to handle time-varying resource requirements and availabilities, activity ready times, due dates and deadlines, fuzzy activity overlaps, activity start time constraints and other types of temporal constraints.
In this study, a method for recognition of Tibetan person name in text based on deep learning is proposed. We presented a better trained Tibetan word embedding and a well-designed detection method for evaluating Tibet...
详细信息
ISBN:
(纸本)9789813146969
In this study, a method for recognition of Tibetan person name in text based on deep learning is proposed. We presented a better trained Tibetan word embedding and a well-designed detection method for evaluating Tibetan word embedding combined with characteristics of Tibetan. Furthermore, we tested the effectiveness of the proposed method by changing the window size, the number of iterations, and the number of hidden layer units in the neural network.
In this paper, we propose a new data based model for influence maximization in online social networks. We use the theory of belief functions to overcome the data imperfection problem. Besides, the proposed model searc...
详细信息
ISBN:
(纸本)9789813146969
In this paper, we propose a new data based model for influence maximization in online social networks. We use the theory of belief functions to overcome the data imperfection problem. Besides, the proposed model searches to detect influencer users that adopt a positive opinion about the product, the idea, etc, to be propagated. Moreover, we present some experiments to show the performance of our model.
The workflows of industrial and design enterprises are represented by diagrams in notation languages: Business Process Model Language (BPML), Unified Model Language (UML), and Architecture of Integrated information Sy...
详细信息
ISBN:
(纸本)9789813146969
The workflows of industrial and design enterprises are represented by diagrams in notation languages: Business Process Model Language (BPML), Unified Model Language (UML), and Architecture of Integrated information Systems (ARIS), Methodology Integration Definition (IDEF0 and IDEF3). To analyze, control and correct made by a designer during development errors in diagrams, the authors have developed a new temporal grammar called TRVM-grammar. It allows the authors to increase quality control and compatibility of the flows of the project works by 20% in comparison with the competing use of preserving, positional and relational grammars.
Improving the efficiency of computer-based training systems is connected with their intellectualization. Research and development of intelligent learning environments for organization of the traditional University (sc...
详细信息
ISBN:
(纸本)9789813146969
Improving the efficiency of computer-based training systems is connected with their intellectualization. Research and development of intelligent learning environments for organization of the traditional University (school) and corporate (within corporations and enterprises) education is an urgent problem. In the research work authors analyzed adaptive learning systems and identified their shortcomings. The authors proposed an approach, principles, models, architecture and tools for the intelligent learning environment that provide personalized training on the basis of dynamically knowledge change. Important attention was paid to the development of virtual component of the environment - a system of virtual simulators. The method of analysis of student's work with simulators and CAD systems were proposed.
暂无评论