Learning the structure of a Bayesian network classifier (BNC) encodes conditional independence assumption between attributes. One major approach to mitigate BNCs's primary weakness (the attributes independence ass...
详细信息
ISBN:
(纸本)9789814619967
Learning the structure of a Bayesian network classifier (BNC) encodes conditional independence assumption between attributes. One major approach to mitigate BNCs's primary weakness (the attributes independence assumption) is the locally weighted approach. And this type of approach has been proved to achieve good performance for NB, a BNC with simple structure. However, we do not know whether or how effective it works for improving the performance of the complex BNCs. In this paper, we carry out an systematically experimental analysis to investigate the effectiveness of locally weighted method for complex BNCs measured by the area under the ROC curve ranking (AUC). Experiments on 36 benchmark data sets in IATeka., system demonstrate that although locally weighting significantly improve the performance of NB (a BNC with simple structure), it could not work well on BNCs with complex structures.
Ensemble of classifiers is a very popular method for online and incremental learning in non-stationary environment, as it improves the accuracy of single classifiers and is able to recover from drifting concept withou...
详细信息
ISBN:
(纸本)9789814619967
Ensemble of classifiers is a very popular method for online and incremental learning in non-stationary environment, as it improves the accuracy of single classifiers and is able to recover from drifting concept without explicit drift detection. However, current ensemble weighing methods do not consider the relationship between a test instance and each ensemble member's training domain. As a result, a locally correct ensemble member may be reduced weight unfairly because that its prediction result of an out of domain test instance is wrong. These inaccuracies will increases when there is a significant concept change. In this paper, therefore, we proposed a fuzzy online ensemble weighting method which takes the consideration of the degree of membership of each instance in each ensemble member and a modified majority voting method to improve the ability of ensemble on handling online classification tasks with concept drift.
This paper presents a new architecture for a Spatio-Temporal Decision Support System. It was designed taking into account epidemiological aspects for decision making in public health management. The main goal is analy...
详细信息
ISBN:
(纸本)9789814619967
This paper presents a new architecture for a Spatio-Temporal Decision Support System. It was designed taking into account epidemiological aspects for decision making in public health management. The main goal is analyze the spatial and spatio-temporal features of a geographic area and make decisions about prevention and control of a disease dissemination on that area. An example of application of this architecture is presented to construct a system for real aids data is presented, as well as the results obtained from it.
In this paper, we studied the classes of quasi-homogeneous overlap functions. In which we demonstrate that all the class of quasi-homogeneous overlap functions properly contains the class of quasi-homogeneous t-norms,...
详细信息
ISBN:
(纸本)9789814619967
In this paper, we studied the classes of quasi-homogeneous overlap functions. In which we demonstrate that all the class of quasi-homogeneous overlap functions properly contains the class of quasi-homogeneous t-norms, that is, all quasi -homogeneous t-norm is a quasi-homogeneous overlap function, but not all overlap function quasi-homogeneous is a quasi-homogeneous t-norm.
Today, most companies monitor the satisfaction of their employees. Companies are trying to measure the satisfaction of their employees and with this aspect they are making efforts to increase workload and work quality...
详细信息
ISBN:
(纸本)9789814619967
Today, most companies monitor the satisfaction of their employees. Companies are trying to measure the satisfaction of their employees and with this aspect they are making efforts to increase workload and work quality, which is beneficial for the company itself. This article proposes a fuzzy tool for assessing the employee's satisfaction within the company. A part of the fuzzy expert system is a tool for evaluating satisfaction questionnaires. All the parts of the system are introduced and verified on a specific example.
Domain adaptation addresses the problem of how to utilize a model trained in the source domain to make predictions for target domain when the distribution between two domains differs substantially and labeled data in ...
详细信息
ISBN:
(纸本)9789814619967
Domain adaptation addresses the problem of how to utilize a model trained in the source domain to make predictions for target domain when the distribution between two domains differs substantially and labeled data in target domain is costly to collect for retraining. Existed studies are incapable to handle the issue of information granularity, in this paper, we propose a new fuzzy domain adaptation method based on self-constructing fuzzy neural network. This approach models the transferred knowledge supporting the development of the current models granularly in the form of fuzzy sets and adapts the knowledge using fuzzy similarity measure to reduce prediction error in the target domain.
New trends in recommender systems face new challenges as group recommendation, in which users give their preferences over items and the system provides recommendations for a group of known users. In certain types of g...
详细信息
ISBN:
(纸本)9789814619967
New trends in recommender systems face new challenges as group recommendation, in which users give their preferences over items and the system provides recommendations for a group of known users. In certain types of groups, it often occurs that several members do not agree on their preferences over some items so their inclusion in the group recommender system (GRS) may mislead the recommendation results. In this contribution a technique to detect and filter conflictive ratings before their use in the recommendation process is proposed and then its performance evaluated by using a well known recommendation dataset. The results show that rating filtering leads to improvements on GRSs performance.
The resources and possibilities to reproduce objective information through infography methods are numberless. This article presents interactive infography contributions to represent data as a result of a systematic ma...
详细信息
ISBN:
(纸本)9789814619967
The resources and possibilities to reproduce objective information through infography methods are numberless. This article presents interactive infography contributions to represent data as a result of a systematic mapping in the area of online social networks analysis. We tried to evidence how interactive infography resources can synthetize complex information in a simple visual representation.
Boolean networks are models of complex dynamical systems. Modelling complex systems with Boolean networks is adequate in situations in which binary view is valid. If the situation is not "black and white", B...
详细信息
ISBN:
(纸本)9789814619967
Boolean networks are models of complex dynamical systems. Modelling complex systems with Boolean networks is adequate in situations in which binary view is valid. If the situation is not "black and white", Boolean networks are inadequate/oversimplified models. For modelling real world complex systems, Boolean networks lack descriptive power - fuzzification of the model is required. Boolean networks are fuzzified using the interpolative Boolean algebra. Fuzzy model keeps the Boolean frame.
Computation of approximation in Dominance-based Rough Sets Approach (DRSA) is a necessary step for multi-criteria decision analysis and other related works. This paper presents a parallel approach for computing approx...
详细信息
ISBN:
(纸本)9789814619967
Computation of approximation in Dominance-based Rough Sets Approach (DRSA) is a necessary step for multi-criteria decision analysis and other related works. This paper presents a parallel approach for computing approximations of DRSA. Its feasibility is validated by a numerical example in this paper.
暂无评论