Network information criterion-determining the number of hiddenunits for an artificial neural network model

The problem of model selection, or determination of the number of hidden units, can be approached statistically, by generalizing Akaike's information criterion (AIC) to be applicable to unfaithful (i.e., unrealizable) models with general loss criteria including regularization terms. The relation between the training error and the generalization error is studied in terms of the number of the training examples and the complexity of a network which reduces to the number of parameters in the ordinary statistical theory of AIC. This relation leads to a new network information criterion which is useful for selecting the optimal network model based on a given training set.     

H∞ control versus disturbance-observer-basedcontrol

In the motion control field, a disturbance observer-based disturbance canceling control is often used as a robust control methodology. However, this method is nothing more than an alternative design of an integral controller, and the robust stability issue cannot be directly accounted for. In this paper, an extended H_∞ control scheme is proposed as a new robust motion control method which achieves the disturbance cancellation ability and guarantees robust stability automatically     

Visualizing the structure of Web communities based on data acquired from a search engine

Discovery of Web communities, groups of Web pages sharing common interests, is important for assisting users' information retrieval from the Web. This paper describes a method for visualizing Web communities and their internal structures. visualization of Web communities in the form of graphs enables users to access related pages easily, and it often reflects the characteristics of the Web communities. Since related Web pages are often co-referred from the same Web page, the number of co-occurrences of references in a search engine is used for measuring the relation among pages. Two URLs are given to a search engine as keywords, and the value of the number of pages searched from both URLs divided by the number of pages searched from either URL, which is called the Jaccard coefficient, is calculated as the criteria for evaluating the relation between the two URLs. The value is used for determining the length of an edge in a graph so that vertices of related pages will be located close to each other. Our visualization system based on the method succeeds in clarifying various genres of Web communities, although the system does not interpret the contents of the pages. The method of calculating the Jaccard coefficient is easily processed by computer systems, and it is suitable for visualization using the data acquired from a search engine.     

Link Prediction based on Structural Properties of Online Social Networks

Question-Answering Bulletin Boards (QABB), such as Yahoo! Answers and Windows Live QnA, are gaining popularity recently. Questions are submitted on QABB and let somebody in the internet answer them. Communications on QABB connect users, and the overall connections can be regarded as a social network. If the evolution of social networks can be predicted, it is quite useful for encouraging communications among users. Link prediction on QABB can be used for recommendation to potential answerers. Previous approaches for link prediction based on structural properties do not take weights of links into account. This paper describes an improved method for predicting links based on weighted proximity measures of social networks. The method is based on an assumption that proximities between nodes can be estimated better by using both graph proximity measures and the weights of existing links in a social network. In order to show the effectiveness of our method, the data of Yahoo! Chiebukuro (Japanese Yahoo! Answers) are used for our experiments. The results show that our method outperforms previous approaches, especially when target social networks are sufficiently dense.

Development of an epileptic seizure prediction algorithm using R–R intervals with self-attentive autoencoder

Epilepsy is a neurological disorder that may affect the autonomic nervous system (ANS) from 15 to 20 min before seizure onset, and disturbances of ANS affect R–R intervals (RRI) on an electrocardiogram (ECG). This study aims to develop a machine learning algorithm for predicting focal epileptic seizures by monitoring R–R interval (RRI) data in real time. The developed algorithm adopts a self-attentive autoencoder (SA-AE), which is a neural network for time-series data. The results of applying the developed seizure prediction algorithm to clinical data demonstrated that it functioned well in most patients; however, false positives (FPs) occurred in specific participants. In a future work, we will investigate the causes of FPs and optimize the developing seizure prediction algorithm to further improve performance using newly added clinical data.

     

Negative emotion recognition using multimodal physiological signals for advanced driver assistance systems

Artificial Life and Robotics - Recent advanced driver assistance systems’ (ADASs) control cars to avoid accidents, but few of them consider driver’s comfort. To realize comfortable...