A front and back simultaneous display of the spherical SOM was newly developed. This makes it possible to display very clearly that the phase indication on the back side of the spherical surface was unknown up to now. So that, the phase relation becmomes now clearer. Detailed results of application of simple animal map, medical health care data, and gene analysis data with this tool will be described in detail. In addition, we applied it to genetic data with a large amount of label this time, and succeeded in raising the resolution by increasing the number of surface maximum nodes of 2562 to 10242 at present.

The purpose of this study was to investigate the toe-distance and heel-distance for young people stepping over an obstacle on a dual-task condition. Subject characteristics were as follows: age range 21-22 years, height range 154-180 cm, and step range 31-67 cm. The experiment was performed in a large room with a straight 8 m travel path available for the subjects to walk along. Test conditions were as follows: double obstacles at 2cm and 2cm (C1), 5cm and 5cm (C2), 10cm and 10cm (C3), 30cm and 30cm (C4), 2cm and 10cm (C5), 2cm and 30cm (C6), 10cm and 2cm (C7), and 30cm and 2cm (C8) positions. The distance between the obstacles were 2steps of each subject. As a result of this study, step over the motion of dual task has the characteristics of toe-distance and heel distance. Therefore, it was shown that second obstacle has influenced the toe-distance and heel distance.

This paper discusses the effects of listening to ultrasonic sound with 1 / f fluctuation in the autonomic nervous system by conducting two experiments. The first is an experiment that increased the number of subjects from the previous study. The second is an experiment to confirm what kind of 1 / f fluctuation can predominate parasympathetic activity. As a result, in the first experiment, it was confirmed that subjects with active parasympathetic activity existed when 1 / f fluctuation sound source was given. In the second experiment, it was confirmed that subjects with more active parasympathetic nervous activity existed as the degree of modulation increased between 10% and 30%. From these results, this study discovered unique effect of 1 / f fluctuation for the autonomic nervous system.

In the sentiment analysis, we usually classify costumer reviews into some classes based on their topics. Technical term dictionary is very important to classify them into correct classes by some classifier such as support vector machine (SVM) learned by the customer review data. In other previous researches, the technical term dictionaries have been constructed by high frequency words in the same topic data. In this paper, we propose a method to construct a high quality technical term dictionary by extracting relevant words and their distributed representation for each topic semi-automatically and we estimated an effectiveness of our proposed method using experimental results.

The contingency table analysis provides the interrelation between two variables and can help find interactions between them. Previously, we extended the contingency table by applying fuzzy theory and defined a fuzzy contingency table (type-1 fuzzy contingency table).Further, we extend type-1 fuzzy contingency table to type-2 in order to analyze inexact information. In this paper, we apply the type-2 fuzzy contingency table to similarity index calculation.．

The purpose of this paper is to investigate the quantitative changes of autonomic nerve activity using AC electromagnetic fields. In previous research, we attempted to measure autonomic nervous activity using a Heart Rate Variability (HRV). However, due to influence of the magnetism, a correct result could not be obtained. In this research, we measured the autonomic nerve activity not affected by magnetism using measuring equipment of skin potential activity. The experiment was repeated for one subject. As a result, it was found that the skin potential level tends to rise and skin potential reflex fluctuates when magnetic exposure and after the exposure. It is thought that the blood flow was accelerated by magnetism, and the sympathetic nerve activity became dominant. However, this experiment haven’t had examined enough data regarding the researched subjects. For this reason, there is a need of gathering more data for the classification of autonomic nerve activity.

In this paper, we describe a discussion on the facilitation of work in the school library using BGM (Back Ground Music) such as classical music, pops and jazz, etc.. We conducted experiments assuming two kinds of work. That is reading work and calculating work. In the survey by questionnaire after these works, we obtain an evaluation that it is easier to work when there is BGM than when there is not BGM. However, even in the presence of BGM, we could not confirm the tendency of the amount of reading and calculating work to improve. This is presumed that because the presentation time of BGM was short, subjects concentrated on the presented BGM, which affected work. In the future, we will reconsider the presentation time of BGM, and we will further investigate.

We have investigated effects of hearing the 1/f fluctuation sound source in ultrasonic range on the automatic nerve activity (the parasympathetic nervous system). At the moment, it show the possibly that 1/f fluctuation sound in ultrasonic range has some effect on the automatic nerve activity. However, it is yet to reveal the parameter of the 1/f fluctuation signal sequence (period, fluctuation width, modulation method, etc.,) that have effect of making the parasympathetic nerve activity predominant. In this study, we investigated the relationship between the modulation method (the frequency modulation) and the parasympathetic nerve activity. As the result, the 1/f fluctuation sound source in ultrasonic range generated by the frequency modulation cannot swear to make the parasympathetic nerve activity predominant.

In sentiment analysis of travelers’ reviews, it is important to understand under what conditions the opinions were written by reviewers. We studied an estimation method of descriptions of estimation condition in travelers’ reviews using conditional random fields. In the previous research, 8 features in textual data were selected for conditional random fields. In this research, we proposed 2 new features and verified effectiveness of the proposed features by experiments. As the result, the F-measure was increased from 0.798 to 0.829.

This paper proposes a method for estimating similarity of recipe sentences employing distributed representation and Word Mover’s Distance (WMD). Recipe sharing service on internet is increasingly popular and have huge amount of recipe. Shared recipes have tags and other information, and we can easily categorize recipes by using these tag information. But tags attached to recipes are sometimes incorrect or not enough. So estimating distance (reverse of similarity) of recipes by using recipe sentence is useful to categorization and bring benefits to user. In this research, we investigated effectiveness of our method.

Serum uric acid level is considered to be a useful index related to metabolic syndrome. We examined the classification by using the SOM significance degree method and compared it with the result of multivariate discriminant analysis. In this study Kohonen's LVQ (Learning vector quantization) was applied to the classification. The result was reported and compared with multivariate discriminant analysis.

A trial of “Bread recognition by the r-θ histogram ” was done and the following results are obtained．1.Changing the orientation or the location of the bread gives rise up the complicated change of the histogram. 2.When the origin of the coordinate system is taken at the center of the bread region, the histogram becomes about identical except the shift along the θ axis. 3.The shift along the θ axis will be disappear when the original histogram is shifted as the maximum histogram angle becomes to the zero.

When analyzing uncertain data expressed by a set of data points, calculating the density difference between any two sets of data points is an important task. Least squares density difference (LSDD) estimator which uses a non-parametric model approximates the true density difference accurately. In the previous method, it searches for the parameters of the model by using the grid search. However, there is no guarantee that the obtained parameters by the grid search are optimal, we apply a firefly algorithm which is suitable to the multimodal optimization problem for the parameter search. Numerical experiments are conducted to show that the LSDD estimator by using the firefly algorithm approximate the true density differences and L²-distances more accurately than the grid search.

Radiation transport simulation codes are used to make an assessment of proton and heavy ion radiotherapies and to make a calculation of Radiation shield for medical accelerator facilities. Double differential cross section (DDX) of charge particle production is one of the important elements to assess and improve such simulation codes. We investigated Electric charge particle production DDX using 100MeV/u alpha particle beam on aluminum and cobalt targets. Charge particle DDX haven’t been reported for alpha of 100MeV/u incident reaction. Emitted charge particle’s energies were measured with spectrometer consisting of NE102 detectors, BSO crystal scintillation detector, and PWO crystal scintillation detector by the ΔE-E technique. The experimental data of DDX were compared with radiation transport simulation code PHITS.

This paper presents how music influences portrait drawing. We had participants listen to an identical melody at slow (75 bpm), medium (100 bpm) or fast (150 bpm) tempos as they drew a portrait for 7 minutes while we used an eye tracking system to acquire their gaze point coordinates. The results of analyzing the GTF (glance translation frequency a minute between the drawing area and the portrait model), the fast tempo group's frequency increased more than the slow tempo group's frequency. One-way analyses of variance with 3 levels of musical tempos were performed on the GTF data. There was a significant difference between the fast tempo and the slow tempo. This suggests that faster music tempos may improve observational abilities and concentration levels.

We developed a new piece of software capable of simultaneously displaying both sides of the spherical SOM. This enables simultaneous comparison, making it much easier to determine the correspondence between the back and front. Moreover, by making a modified the software to 64-bit version, it also supports for large-scale data. Furthermore, an implementation of high-speed processing has been made, and the evaluation of it was carried out. And on the newly developed software, the maximum number of node for Spherical SOM is increased 2562 to 10242.

We developed an original observation system and measurement equipment of the Raleigh-Benard model (aspect ratio 2.17) in a solenoidal superconducting magnet to visualize magneto-thermal convection in situ from the side view. The working fluid of convection was an aqueous gadolinium nitrate solution of 0.15 mol/kg (paramagnetic substance), and the isothermal contour of convection was visualized by the addition of thermosensitive liquid crystal (TLC). We succeeded in visualizing the horizontal isothermal illuminant of TLC,which revealed the realization of a quasi-weightless condition in the Raleigh-Benard model. The volume of heat transfer in this system was measured, and then the Rayleigh number (Ra) and Nusselt number (Nu) were experimentally determined. We identified that the heat transfer on convection was enhanced by the downward magnetic force and suppressed by the upward magnetic force, as compared with Raleigh-Benard convection.

this study aims to classify the shape of maxillofacial skeleton and find relationships between maxillofacial skeleton and swallowing function．We used SOM² for the shape classification and CCA-SOM for search the relationships．Previously，We use shape address for CCA-SOM．But We tried to use SOM²’s unit for CCA-SOM．Results show that We cannot find the relationships by using both methods．Next plan is execute CCA-SOM by using less-dimension maxillofacial skeleton data.

As a method of detecting CTC, there is a method of measuring angular velocity by electrorotation. However, the angular velocity varies greatly even for cells of the same type. In this paper, assuming that the shape is a cause of the cause, we investigate the correlation between cell shape and angular velocity. Two methods were used for shape representation. One is a method of expressing using area and perimeter. The other is to map the distance set data between the center of gravity and the contour on the two-dimensional plane using the SOM. Also, the angular velocity is corrected based on the shape.

Dendrogram is used for cluster classification. Using this classification, the spherical surface was colored with the clusters. Thus, the misclassification becomes clearer on the sphericalsurface. There was no misclassification in vital data for 25 people. However, with 399 people vitaldata, there were misclassification of 8 labels. The misclassification of the 7 labels could beunderstood by distance using these input data. For the remaining one, the distance was computedusing the codebook vector of learned scod. In this way, the reason for the misclassificationbecame clearer.

For elderly people with weak walking ability, it’s difficult to participate in various social activities, and become healthier with spirit. In this study, novel shoes which can sense the gradient of the ground are developed for easily walking on slopes. These shoes can adjust the position of heel and tiptoe according to the gradient of road. In the report, we investigated the load placed on the legs when ascending a slope with a heel height adjustable shoes. In particular, walking experiments were conducted on a treadmill machine with gradients ranging from 0% to 20%. The activities of leg muscles was measured by wireless myoelectric sensors to clarify the effectiveness of the heel altitude adjustment.

Routine health care is necessary to prevent lifestyle-related diseases that have a high proportion as death cause of the Japanese. At present, the use of plethysmogram as a means of health care is noticed. And the plethysmogram can measure arteriosclerosis which is the cause of lifestyle-related disease. However, the measurement of arterial stiffness has a problem that the measurement cannot be carried out unless it goes to a specialized medical institution, and the time and action are restricted during measurement. The conventional plethysmogram measurement is performed using red light, but in recent years, it is noticed that the plethysmogram measurement of green light is better. In order to solve these problems, we will manufacture a wearable arterial stiffness measurement system using green LEDs.

The knowledge of various ways to achieve the same purpose of program is important for program education and software engineering. In this paper, we propose a vector representation of program with that the structure of abstract syntax tree of a program is appropriately indicated. In order to show usefulness of our proposed method, we conduct the following experiments: firstly, we confirm that our proposal can analyze fundamental programs as appears in textbooks based on the representation. Secondly, we also confirm that our proposed method can classify ways of programing that are adopted in various programs submitted to an assignment of program. From these results, we conclude that our proposed method is a useful for analyzing how a program achieves its purpose.

Autoencoders that acquire specific feature space models from unsupervised data have become an important technique for designing systems based on neural networks. In this paper, we investigate of the features of high reusable autoencoder. We confirmed this ability from two experiments, focusing on the role reusable auto encoders. As a result, we found that the role that the intermediate units of the most reusable autoenocder in our experiments plays is regarded as binary function.

In this paper, we assess models to distinguish the styles of writing in three different fields. The models are trained with the neural network and we employ the occurrence frequency of the end-of-sentence expressions in each document as feature to machine learning. For evaluating adequacy of each trained model to target problem, we compare the complexity of each trained model with AIC. Furthermore, showing the distribution of the examples on reduced variable space with the trained model, we confirmed its outcome as information compression from the original feature space.

Precious opal is a gemstone consist of multiple regions where the light is scattered or diffracted. We have proposed a method for modeling this structure of opal using volumetric representation, and a rendering method based on the photon mapping algorithm. This time, the absorption has been also considered. As the result, we can successfully reproduce a variety of ground colors of opal.

Recently, AI technology called deep learning has been attracting attention. Deep learning is a kind of neural network. It is a method of machine learning using a hierarchical neural network with a multi-layer structure. Deep learning has shown great results in pattern recognition. In this study, we consider the classification of unearthed coins using deep learning. There are many types of unearthed coins to be excavated from a ruin. However, the classification of unearthed coins is even difficult task for numismatics experts. The objective of this research is to conduct basic studies into the classification of unearthed coins by deep learning.

Many studies have been done with security of cloud computing. Secure Multiparty Computation (SMC) is one of these methods. In previous papers, some learning methods with SMC have been proposed such as Back Propagation, k-means and digital Q-learning, but analog Q-learning for SMC has not been proposed yet. In this paper, we propose an analog Q-learning method for SMC and show the effectiveness by the simulation.

A monitoring system for the elderly by using a Laser Range Finder (LRF) is being developed. The long-term objective of the system is to extract daily life patterns of an elderly person based on trajectory data and to detect an anomaly if, when observing a new life pattern, it significantly differs from the daily life patterns. A life pattern represents a series of behavior transitions such as Toilet => Washstand => Closet => Bed. In such places, it is assumed that an elderly person stays for a certain time of period. In this paper, such places are called “stay points”. This research focuses on those methods that automatically estimate the stay points from movement data. Clustering approach is employed for this purpose. Numerical experiments were conducted by considering situations that may occur in elderly facilities. Through a series of computational experiments, we investigate which combination of clustering method and evaluation index provides the best estimation of the stay points.

In previous studies, genuinely silent sound environment was created using sound masking at a Polytechnic College library. In the research, one user commented “I can concentrate better when there is music that I like being played rather than being in a completely silent environment”. From these results, we suggest that the possibility is as follows. Sound environment that is “comfortable” for users improves their concentration and eventually promotes educational and research effects in library utilization. We conducted the preliminary investigation on recognition of Back Ground Music (BGM) in library. As a result, we confirm the following trends regarding recognition of BGM in college libraries.

・ BGM has the effect of enhancing concentration and relaxation for personal learning and reading.

・ Music focused on melody and harmony is preferable as BGM, whereas music focused on vocals is dispreferred.

Based on these results, we conducted an auditory experiment on BGM. In the experiments, users performed reading and calculation tasks in the library. The results showed that subjects found reading and numerical tasks easier to perform when there is BGM. In addition, no clear correlation was found between BGM that subjects found easier to work with and their personal taste in music. In this paper, we will discuss the preliminary survey described above and the auditory test. Subsequently, we will discuss developing a sound environment in a library that heightens users’ power of focus and promotes work.

As low birthrate and the aging of the population progresses, labor force is seriously lacking in heavily worked construction sites including material handling recently because it is very difficult to replenish young people according to retirement elderly number. If we can leave material handling work to the robot, the problem of labor shortage will be relaxed. In this research, we aim to develop a two wheels independent driven robot that transports materials, and developed a path following control method by the PID control method, and conducted a simulation to show its effectiveness.

A general password authentication method can’t prevent unauthorized use by a third party if a system would be hacked (cracked) in some way even once. Therefore, we propose a continuous personal authentication method with keystroke data for server operation which is required at that situation. If server administrators would be requested to input their own password periodically again and again, the users will be too busy to operate system and the burden will increase absolutely. Therefore, it can be said that it is important to use a method of personal authentication that does not require user consciousness. So in this research, in order to alleviate the burden on users and solve security problems, we verify whether personal authentication by self-organizing map is possible using keystrokes during server operation and also evaluate the availability of our proposed method.

Medical clerk was introduced in 2008 to reduce the burden on doctors working in hospitals. A medical clerk refers to a medical staff that assists or acts office work such as preparing a medical certificate made by a doctor. In this paper, at the viewpoint of medical safety, we analyze and consider the risks and countermeasures for medical safety due to the intervention of medical clerk with the risk probability model.

The effects of treadmill running impact on the runner’s fatigue state were often examined from interactions among corresponding attributes. While fatigue accompanied by tired legs, muscle pain and cramps were often described, no study has classified the levels of treadmill running motion into fatigue and non-fatigue conditions using data mining approach. This study identifies the significant attributes that promote accurate classification of treadmill running fatigue conditions. The study data was retrieved from a benchmark public database that exhibits treadmill running variability. The qualitative and quantitative features of anthropometric, demographic, accelerometry and sacral trajectory attributes were initially evaluated for data classification. Five classifiers: Lazy, Function, Meta, Rules and Tree built-in Waikato Environment for Knowledge Analysis (WEKA) tool were employed on the training and 10 folds cross validation test modes. The [CorrelationAttributeEval] and [WrapperSubsetEval] were applied for attribute evaluation in order to enhance classification accuracies at 10 folds cross validation mode. This study distinguishes significant attributing data features into fatigue and non-fatigue levels on average classification accuracies for the before and after significant attribute considerations: Lazy (28.3%, 63.3%), Function (56.4%, 59.3%), Meta (45.9%, 60.0%), Rules (46.5%, 59.0%) and Trees (48.2%, 65.0%). Findings also reveal that the RMS of mediolateral is the most significant attribute whose variations exert a major effect on the treadmill running fatigue classes.

The vitreous is a transparent gel-like tissue filling the eyeball. It sometimes becomes turbid, and this state is called vitreous opacity. The vitreous opacity is examined through lens. The intraocular lens is an artificial lens, and there are two types of lens. In this study, we evaluate vitreous opacity for each type of lens. In the proposed method, the frames for evaluation are firstly selected and the evaluation area is detected. Secondly, the opacities are detected using frame subtraction method, and the degree of vitreous opacity is calculated. In the experiments, the proposed method is applied to the actual image, and the results are verified for each type of lens.

We investigate cerebral blood flow data measured by light topography with jump diffusion model. Since the cerebral blood flow data include los of jumps, it is difficult o find real jump-times from the data. In his paper we consider how to separate jump-times from the observed times.

Because of dementia, patients cannot control excrement activities, and it could cause stress for caregivers in cleaning excrement through gloves or towel paper. Therefore, in this research, we develop a cleaning machine that can clean excrement directly adhering to the floor easily. As the first step of this research, a mechanism can clean the floor with a highly viscous substance is necessary. Authors propose an excrement cleaning machine with 3 cleaning parts, brush or paper towel for dropping dirt directly, a device for spreading liquid, and a device for vacuuming dirt. In this report, we investigated cleaning performance of the proposed prototype of the excrement cleaning machine by different combinations each cleaning parts.

This paper proposes a new concept of subjective evaluation measurement based on the Visual Analog Scale (VAS). The purpose of this concept is the visualization of the sobering of subjective evaluation. And it is possible that this concept grasps the overall of the subjective evaluation. In order to realize these, this concept consists the following elements; (1)Measurement of subjective evaluation based on VAS, (2)Calculation of descriptive statistics value of the obtained subjective evaluation value, (3)Creating a box-and-whisker plot and (4) Implementation of the cluster analysis (ward method and k-means method). Under the above preparation, we verify the effectiveness of this new concept by actual subjective evaluation experiment.

This paper describes the relation between infantile crying and thrombus generation near intravenous indwelling needle in veins. In this study, it is assumed that the difference of pressure or flow rate between normal and infantile crying causes thrombus formation. First of all, the blood pressure of infant are measured, and the difference of blood pressure between normal status and infantile crying. It was found that the blood pressure increases according to the infantile crying, and the difference of blood is approximately 20 mmHg. To investigate the flow field of the flow, CFD analysis around the indwelling needle of drip injection (catheter) in the infantile vein was done. The results show that the increasing flow rate or velocity make the accumulation of particle around the inlet of the catheter.

In the previous study, a standing-up assistance method of the walking-support robot with less burden on users’ body was developed according to acceleration information of user and the effectiveness was verified by experiments on healthy subjects. In this report, for verifying the standing-up assistance performance for the weak elderly of the proposed method, standing-up assistance experiments by the walking-support robot was performed with a pseudo weak elderly who was restrained on the knee. The effectiveness of the standing-up assistance method was analyzed in this research.

In order to perform walk rehabilitation for the patients who cannot maintain their standing postures eaely, seat-style walk training robot is under development in our laboratory. By this robot, trainees can make walk rehabilitation in the sitting posture. However, the rehabilitation performance of this robot was not evaluated in quantity.

In this study, we proposed an evaluation method of rehabilitation performance by measuring the sEMG of lower limb muscle during walking in sitting position. Specifically, the sEMG signal of muscular for walking forth, back, right and left by changing the seat height was measured in this research. The usefulness walk training by the seat-style walk training robot was verified.

In recent years, rescue robots have been gradually developed. For rescue robot, it is required to be able to travel on rough terrain without falling down. Therefore, in our laboratory, we are conducting research on omnidirectional a mobile robot which can get rid of the risk of falling upside down. It is proposed that a novel wheel can run in all direction and rough terrain. In this study, we explain a design and mechanism about it and verify the effect of it by running experiments on rough terrain.

In this study, we evaluated the shock mitigation armor at the time of falling backwards, and as a result we found that armor protecting head and neck is the best result. Moreover, even if the shock was relieved, it was found that the armor where the head touches the ground is not suitable for a rearward fall

We discuss an essence of Arrow's impossibility theorem and the effectiveness of fuzzy vote for the easy evasion of the paradox in collective choice theory.

In this paper, we examined to reveal the relationship between aromatherapy and automatic nerve activity quantitatively. Tanita reported that parasympathetic activity becomes dominant in subjects who like its fragrance and sympathetic activity becomes dominant in subjects who don't like its fragrance. However, those subjects were only female. Therefore, our purpose is to verify for whether men can achieve the same results as women and to considerate correlation between subjective evaluation by SD method and objective evaluation by LF/HF. In this experiment, we measured LF/HF for 5 minutes each in rest state and fragrance inhalation state, we measured the change in emotions before and after the experiment,and we evaluate preference of the essential oil. As a result, we obtain a tendency that parasympathetic activity becomes dominant in subjects who like its fragrance and sympathetic activity becomes dominant in subjects who don't like its fragrance, same as Tanita's report. Therefore it is clear that there is no gender difference in the autonomic nervous activity. Furthermore, changes in autonomic nervous activity and emotions changes are consistent.