Sleep disorders are modern diseases that have recently increased in prevalence. In particular, the rapid-eye-movement (REM) sleep behavior disorder (RBD) is well-known. As a hallmark of the RBD, we can find the loss of skeletal muscle atonia during the REM sleep, which is called the REM sleep without atonia (RWA), in the polysomnography (PSG). Therefore, measurement of muscle activity is required for the diagnosis of RBD. Although close follow-up of patients' recovery process is required and several visual inspection methods have been developed for the PSG reading, clear diagnostic criteria for assessing RBD severity have not been established because of the complicated nature of the procedure. In this study, the author constructed automated algorithms based on the AASM (American Academy of Sleep Medicine) scoring manual. The percentage ratio of RWA in the REM sleep period was calculated by the algorithms. The results evaluated by the automated algorithms were compared to the diagnosis by the visual inspection method. In addition, nonlinear discriminant analysis was employed to examine whether the healthy subjects group was significantly different from the group of suspected RBD patients. The author has succeeded in finding an automated algorithm to yield results that are not significantly different from those of visual inspection diagnosis.

In recent years, it has been pointed out that the management competitiveness is shifting from the functional value such as high performance and advanced technology to the emotional value such as experience and design. However, there are few cases that quantitatively evaluated it. Therefore, I validated a hypothesis that emotional value is more important than functional value as a factor that contributes to corporate brand favor based on pure recall data.

This paper proposes a dictionary consultation method for Web reading. Natural language expressions appear in texts overlapping linguistic units, such as characters, words, and phrases, which can be represented hierarchically. The proposed method takes hierarchical structures of languages into consideration and provides a way for the reader to select a specific natural language expression from a string and display associated linguistic information. The proposed method simplifies the use of language resources when unknown natural language expressions are encountered. The proposed method has been compared with several conventional methods for input usability. The result of the comparison has confirmed that the proposed method has characteristics that are superior to conventional methods.

Various methods to predict stock prices have been studied. In the field of empirical finance, feature values for prediction include “value” and “momentum”. In this research, we use the pattern of stock price fluctuations which has not been fully utilized in the financial market as the input feature of prediction. We extract the representative price fluctuation patterns with k-Medoids Clustering with Indexing Dynamic Time Warping method. This method is k-medoids clustering on dissimilarity matrix using IDTW which measures DTW distance between indexed time-series. We can visualize and grasp a price fluctuation pattern effective for prediction with the proposed method. To demonstrate the advantages of the proposed method, we analyze its performance using TOPIX. Experimental results show that the proposed method is effective for predicting monthly stock price changes.

Various methods to predict stock prices have been studied. In a linked paper, we propose a stock price prediction method that applies DTW on scaled daily stock price patterns. We showed its performance using TOPIX data. In this paper, we further analyze the effectiveness of the proposed method using other price indexes, such as the S&P 500, CAC40, DAX, FTSE100, foreign exchange rates, and Cryptocurrency. This paper also reports the importance of pattern length and starting points of patterns. We also discuss the hidden behavior of investors which seems to explain the prediction ability of the proposed method.

In this research, we aim to develop a card operation-based programming learning system focusing on thinking between the relations of parts and evaluate its effectiveness from the viewpoint of cognitive load. Programming learners need various abilities such as creativity, mathematics, logical thinking, and structure grasping to write a program. In addition, there are some kinds of cognitive load in the programming learning such as typing, computer operation, mathematical thinking, and algorithm design. We can control cognitive resources by designing learning materials which limit learning activity patterns, and it will be effective for efficiently supporting novice programming learners. Therefore, this paper focuses on the parts consisting of one or more statements in a program and makes learners concentrate on thinking between the relations of the parts. To realize this concept, this paper proposes “card operation-based method”, a method to complete a program by arranging cards. The main target of this paper is college students who are programming beginner. We introduced this system to an actual class and carried out the learning support. As the result, the proposed system was able to focus the learners on the content intended by the instructor while reducing cognitive load.

OSS (Open Source Software) has made remarkable progress in recent years. In ES (Enterprise Software) development process, also the OSS usage has been increasing steadily, which means the impact of OSS to the society has become greater and greater. However, the quantitative analysis for OSS development has been put no particular importance today. In this research, we study to enhance COCOMO model, commonly used for ES development, in order to clarify the relationship between human resources and development volume of OSS development projects. We chose 50 of Large Scale OSS development Projects from GitHub for our analysis, and obtained such three findings as: 1. In OSS development projects, with some arrangement in measuring man-months volume, we found that power formula of COCOMO model can be applied for expressing the relationship between the development volume and the outputs. 2. As OSS development characteristics, with the increase of development volume, the efficiency for development shows the rising trend. 3. The tendency found above has been identified as not the results of diversion as copying and modifying sources nor cumulative contribution made by variety of small developers.

Offloading which means the enhancement of link capacity is performed to accommodate ever-increasing network traffic. We study determination of enhancement links effective even if there are many links showing the high same link utilization ratio. And also, it is desirable that the link enhancement is performed on a stepwise link pair which has the greatest effect of improving congestion and achieves an effect each time it is increased by one. In this paper, first, we propose a determination method of offloading link section and its effectiveness is verified. Second, we propose and discuss a stepwise link decision method that maintains the effective improvement of congestion reduction.

Recently, Wireless Power Transmission (WPT) system in accordance with Qi standards is being installed into smartphones. However, the transmission efficiency greatly deteriorates due to adjacent metal objects such as battery cases. To solve this problem, it is thought that the high permeability magnetic sheet insertion method into the gap between WPT-coil and metal objects is effective. In order to make the thickness of these materials thinner and improve the efficiency, the effective loading method using an amorphous magnetic sheet is proposed in this report.

The authors carried out a physiological evaluation involving plethysmographic analysis and subjective evaluation to identify differences in the usability of personal computer peripherals. As workload items, the speed of the mouse pointer was changed in three stages, and the plethysmography was measured while the users performed a Trail Making Test. Sympathetic nerve activity in the peripheral region and the central trunk region increases when the movement speed of the mouse pointer is slow. However, there was no significant difference in subjective evaluation and activity of autonomic nervous function. Competitive factors of mental workloads suggested that subjects experienced differences in load recognition of psychological stress and that this may affect the activity of autonomic nerve function.

We investigated the correlation and time-lag (LAG) between the sites of the brain for different durations of epileptiform discharges, using wavelet-crosscorrelation (WCC) analysis. Electroencephalography (EEG) was classified into two epochs according to the epileptiform discharge durations, as follows: short run, 2-3 seconds; and long run, 3 seconds or more. Both runs were categorized into before, during, and after the epileptiform discharges. WCC coefficients and LAG in the 6 and 8 Hz bands for each segment were calculated in all the patients. The results of the comparison among all the electrodes in both runs showed that the correlation during the epileptiform discharges was higher than before and after the epileptiform discharges. In the 6 Hz band, in the short run, the results of the comparison among the temporal region electrodes showed that the correlation and LAG between the sites could be stronger in the 2 seconds immediately prior to the beginning of the epileptiform discharges in the EEG. In the long run, the results of the comparison among the temporal region electrodes showed that high correlation and LAG between the sites could be stronger in the 2-4 seconds immediately prior to the beginning of the epileptiform discharges and 2-4 seconds after to the epileptiform discharges in the ipsilateral hemisphere and contralateral hemisphere. It can be suggested that the difference in the duration of the epileptiform discharges may be caused by the balance between the suppressive and excitatory processes in the brain, before and after epileptiform discharges.

Reservoir Computing (RC) is a machine-learning paradigm that is capable to process empirical time-series data. This paradigm is based on a neural network with a fixed hidden layer having a high-dimensional state space, called a reservoir. Reservoirs including time-delays are considered to be good candidates for practical applications because they make hardware realization of the high-dimensional reservoirs simple. Performance of the well-trained RCs depends both on dynamical properties of attractors of the reservoirs and tasks they solve. Therefore, in the conventional monostable RCs, there arise task-wise optimization problems of the reservoirs, which have been solved based on trial and error approaches. In this study, we analyzed the relationship between the dynamical properties of the time-delay reservoir and the performance in terms of the spectra of the delayed dynamical systems, which might facilitate the development of the unified systematic optimization techniques for the time-delay reservoirs. In addition, we propose a novel RC framework that performs well on distinct tasks without the task-wise optimization using bistable reservoir dynamics which can reduce complicated hardware management of the reservoirs.

In this paper, we propose the 1+n/k frequency divider based on the multi-phase clock. This circuit confirms that the desired dividing operation can be obtained under the condition ofn ≤ ⌊k/2⌋ by simulation. Also, since it is a full digital configuration, integration is easy.

Eye movement was measured when food pictures combined with four dishes were subjectively evaluated by seven-grade scale. And the influence of the dish positions to the eye fixation was discussed. Results showed that the eye fixation time in the upper region of pictures was longer than that in the lower region, and the percentage of the region for the first fixation point was larger in the top left among the dish positions.

In this paper an active noise control system using single adaptive filter with adaptive control of the dither gain is proposed. Using the observable signal, the evaluation function and the dither level are calculated independently. As a result, the dither level can be adaptively controlled relating to the evaluation function. Numerical verification of this adaptive algorithm was carried out by computer simulation of the estimation error and the dither level.