• Journal of KIISE:Software and Applications
• /
• v.33 no.10
• /
• pp.877-885
• /
• 2006

Recently it is proposed that the Bayesian networks used as conversational agent for topic inference is useful but the Bayesian networks require much time to model, and the Bayesian networks also have to be modified when the scripts, the database for conversation, are added or modified and this hinders the scalability of the agent. This paper presents a method to improve the scalability of the agent by constructing the Bayesian network from scripts automatically. The proposed method is to model the structure of Bayesian networks hierarchically and to utilize Noisy-OR gate to form the conditional probability distribution table (CPT). Experimental results with ten subjects confirm the usefulness of the proposed method.

• Journal of KIISE:Software and Applications
• /
• v.33 no.3
• /
• pp.344-355
• /
• 2006

In daily activities, the interaction between humans and robots is very important for supporting the user's task effectively. Dialogue may be useful to increase the flexibility and facility of interaction between them. Traditional studies of robots have only dealt with simple queries like commands for interaction, but in real conversation it is more complex and various for using many ways of expression, so people can often omit some words relying on the background knowledge or the context of the discourse. Since the same queries can have various meaning by this reason, it is needed to manage this situation. In this paper we propose a method that uses hierarchical bayesian networks to implement mixed-initiative interaction for managing vagueness of conversation in the service robot. We have verified the usefulness of the proposed method through the simulation of the service robot and usability test.

• Journal of the Korean Statistical Society
• /
• v.31 no.1
• /
• pp.63-75
• /
• 2002

Most attempts at Bayesian analysis of neural networks involve hierarchical modeling. We believe that similar results can be obtained with simpler models that require less computational effort, as long as appropriate restrictions are placed on parameters in order to ensure propriety of posterior distributions. In particular, we adopt a model first introduced by Lee (1999) that utilizes an improper prior for all parameters. Straightforward Gibbs sampling is possible, with the exception of the bias parameters, which are embedded in nonlinear sigmoidal functions. In addition to the problems posed by nonlinearity, direct sampling from the posterior distributions of the bias parameters is compounded due to the duplication of hidden nodes, which is a source of multimodality. In this regard, we focus on sampling from the marginal posterior distribution of the bias parameters with Markov chain Monte Carlo methods that combine traditional Metropolis sampling with a slice sampler described by Neal (1997, 2001). The methods are illustrated with data examples that are largely confined to the analysis of nonparametric regression models.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.1
• /
• pp.72-76
• /
• 2009

Recently, as the interest of ubiquitous computing has been increased there has been lots of research about recognizing human activities to provide services in this environment. Especially, in mobile environment, contrary to the conventional vision based recognition researches, lots of researches are sensor based recognition. In this paper we propose to recognize the user's activity with multi-modal sensors using hierarchical dynamic Bayesian networks. Dynamic Bayesian networks are trained by the OVR(One-Versus-Rest) strategy. The inferring part of this network uses less calculation cost by selecting the activity with the higher percentage of the result of a simpler Bayesian network. For the experiment, we used an accelerometer and a physiological sensor recognizing eight kinds of activities, and as a result of the experiment we gain 97.4% of accuracy recognizing the user's activity.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.12
• /
• pp.5819-5840
• /
• 2018

Considering the topology of hierarchical tree structure, each cluster in WSNs is faced with various attacks launched by malicious nodes, which include network eavesdropping, channel interference and data tampering. The existing intrusion detection algorithm does not take into consideration the resource constraints of cluster heads and sensor nodes. Due to application requirements, sensor nodes in WSNs are deployed with approximately uncorrelated security weights. In our study, a novel and versatile intrusion detection system (IDS) for the optimal defense strategy is primarily introduced. Given the flexibility that wireless communication provides, it is unreasonable to expect malicious nodes will demonstrate a fixed behavior over time. Instead, malicious nodes can dynamically update the attack strategy in response to the IDS in each game stage. Thus, a multi-stage intrusion detection game (MIDG) based on Bayesian rules is proposed. In order to formulate the solution of MIDG, an in-depth analysis on the Bayesian equilibrium is performed iteratively. Depending on the MIDG theoretical analysis, the optimal behaviors of rational attackers and defenders are derived and calculated accurately. The numerical experimental results validate the effectiveness and robustness of the proposed scheme.

• Journal of the Korean Statistical Society
• /
• v.30 no.2
• /
• pp.179-192
• /
• 2001

When a statistical model has a hierarchical structure such as multilayer perceptrons in neural networks or Gaussian mixture density representation, the model includes distribution with unidentifiable parameters when the structure becomes redundant. Since the exact structure is unknown, we need to carry out statistical estimation or learning of parameters in such a model. From the geometrical point of view, distributions specified by unidentifiable parameters become a singular point in the parameter space. The problem has been remarked in many statistical models, and strange behaviors of the likelihood ratio statistics, when the null hypothesis is at a singular point, have been analyzed so far. The present paper studies asymptotic behaviors of the maximum likelihood estimator and the Bayesian predictive estimator, by using a simple cone model, and show that they are completely different from regular statistical models where the Cramer-Rao paradigm holds. At singularities, the Fisher information metric degenerates, implying that the cramer-Rao paradigm does no more hold, and that he classical model selection theory such as AIC and MDL cannot be applied. This paper is a first step to establish a new theory for analyzing the accuracy of estimation or learning at around singularities.

• Journal of KIISE:Software and Applications
• /
• v.41 no.7
• /
• pp.481-491
• /
• 2014

The power production using renewable energy is more important because of a limited amount of fossil fuel and the problem of global warming. A concentrative photovoltaic system comes into the spotlight with high energy production, since the rate of power production using solar energy is proliferated. These systems, however, need to sophisticated tracking methods to give the high power production. In this paper, we propose a hierarchical tracking system using modular Bayesian networks and a naive Bayes classifier. The Bayesian networks can respond flexibly in uncertain situations and can be designed by domain knowledge even when the data are not enough. Bayesian network modules infer the weather states which are classified into nine classes. Then, naive Bayes classifier selects the most effective method considering inferred weather states and the system makes a decision using the rules. We collected real weather data for the experiments and the average accuracy of the proposed method is 93.9%. In addition, comparing the photovoltaic efficiency with the pinhole camera system results in improved performance of about 16.58%.

• Journal of KIISE:Software and Applications
• /
• v.34 no.6
• /
• pp.549-561
• /
• 2007

Vision-based scene understanding is to infer and interpret the context of a scene based on the evidences by analyzing the images. A probabilistic approach using Bayesian networks is actively researched, which is favorable for modeling and inferencing cause-and-effects. However, it is difficult to gather meaningful evidences sufficiently and design the model by human because the real situations are dynamic and uncertain. In this paper, we propose a learning method of Bayesian network that reduces the computational complexity and enhances the accuracy by searching an efficient BN structure in spite of insufficient evidences and training data. This method represents the domain knowledge as ontology and builds an efficient hierarchical BN structure under constraint rules that come from the ontology. To evaluate the proposed method, we have collected 90 images in nine types of circumstances. The result of experiments indicates that the proposed method shows good performance in the uncertain environment in spite of few evidences and it takes less time to learn.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.336-342
• /
• 2001

One of the most important problems on rule induction methods is that they cannot extract rules, which plausibly represent experts decision processes. On one hand, rule induction methods induce probabilistic rules, the description length of which is too short, compared with the experts rules. On the other hand, construction of Bayesian networks generates too lengthy rules. In this paper, the characteristics of experts rules are closely examined and a new approach to extract plausible rules is introduced, which consists of the following three procedures. First, the characterization of decision attributes (given classes) is extracted from databases and the classes are classified into several groups with respect to the characterization. Then, two kinds of sub-rules, characterization rules for each group and discrimination rules for each class in the group are induced. Finally, those two parts are integrated into one rule for each decision attribute. The proposed method was evaluated on a medical database, the experimental results of which show that induced rules correctly represent experts decision processes.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.6A
• /
• pp.550-557
• /
• 2007

We predict a context of various structures by using Bayesian Networks Algorithms, Three-Dimensional Structures Algorithms and Genetic Algorithms. However, these algorithms have unavoidable problems when providing a context-aware service in reality due to a lack of practicality and the delay of process time in real-time environment. As far as context-aware system for specific purpose is concerned, it is very hard to be sure about the accuracy and reliability of prediction. This paper focuses on reasoning and prediction technology which provides a stochastic mechanism for context information by incorporating various context information data. The objective of this paper is to provide optimum services to users by suggesting an intellectual reasoning and prediction based on hierarchical context information. Thus, we propose a design of user location prediction algorithm using sequential matching with n-size flexible window scheme by taking user's habit or behavior into consideration. This algorithm improves average 5.10% than traditional algorithms in the accuracy and reliability of prediction using the Flexible Window Scheme.