We have developed a method of emerging a small-world network, which means the smallest-world network in a sense of absolutely small average path-length compared to other complex networks, in a self-organizing manner using an ACO (Ant-Colony Optimization)-inspired method. We call it an n-Star network. As one of the real-world applications, we showed the n-Star network could be applied to reorganizing a next generation global airline network, where several star nodes are assigned to some of the star cities selected from major cities in the world in advance, and we evaluated the performance of the network using several kinds of network parameters. This method is a hybrid method using a bottom-up and top-down approach. In this study, without selecting any star cities in advance, using a bottom-up method only based on city population, city ranking, and distance between cities, etc., we tried to emerge a self-organizing world airline network by connecting links between important cities. As a result, the latter n-Star network is formed which is different from the former n-Star network, but it is expected that both n-Star networks will concentrate heavy loads to their respective star airports. We will verify the concentration of load to the star airports through a simulation experiment, and propose an effective method for distributing the load over the whole airline network.

The 2011 off the Pacific coast of Tohoku Earthquake on March 11 was a record great natural disaster. To protect us from earthquakes, seismological researchers have been forecasting or predicting earthquakes mainly in the particular areas where earthquake will bring great damages. Some researches point out about the co-occurrence of earthquakes among several areas, however, we have little knowledge about it. In this research, we extract earthquake outbreak patterns of the 2011 Tohoku Earthquake by the proposed method "Co-occurring Cluster Mining", and investigate mechanical correlations among several areas. We acquired novel knowledge about the co-occurrence of earthquake outbreaks.

Distance metric greatly affects the performance of data mining tasks, such as clustering or classification. This paper proposes a distance metric learning based on a global cluster validity measure that evaluates inter- and intra- clusters simultaneously. The proposed method optimizes a distance transform matrix based on Maharanobis distance by utilizing an evolutional algorithm of Differential Evolution (DE). Apart from the most of distance metric learnings, our approach directly improves clustering performance and needs less auxiliary information in principle. In the experiments, we validated the search efficiency of DE, the generalization performance via cross-validation, and also showed how the distance metric learning improves data distribution via visualization by Self-Organizing Map (SOM).

This paper proposes a method to detect faults in artificial heart-lung machines that is used during open-heart surgery to keep blood and oxygen moving throughout the body during the surgery. We use a temporal data mining approach for fault detection and also report the results for past surgeries.

This paper proposes a would-be-worlds browser for socio-economic systems, a new approach integrating a massive-data analysis and agent-based simulation. It was used for workflow analysis and layout design in an operation room. Finally, We discuss the agent-based simulation of multiple persons moving in public spaces and its application to evaluating information presentation for guidance. Using would-be-worlds browser, we can design dependable socio-economic systems.