礒川 悌次郎

Introduction The deep echo state network (Deep-ESN) architecture, which comprises a multi-layered reservoir layer, exhibits superior performance compared to conventional echo state networks (ESNs) owing to the divergent layer-specific time-scale responses in the Deep-ESN. Although researchers have attempted to use experimental trial-and-error grid searches and Bayesian optimization methods to adjust the hyperparameters, suitable guidelines for setting hyperparameters to adjust the time scale of the dynamics in each layer from the perspective of dynamical characteristics have not been established. In this context, we hypothesized that evaluating the dependence of the multi-time-scale dynamical response on the leaking rate as a typical hyperparameter of the time scale in each neuron would help to achieve a guideline for optimizing the hyperparameters of the Deep-ESN. Method First, we set several leaking rates for each layer of the Deep-ESN and performed multi-scale entropy (MSCE) analysis to analyze the impact of the leaking rate on the dynamics in each layer. Second, we performed layer-by-layer cross-correlation analysis between adjacent layers to elucidate the structural mechanisms to enhance the performance. Results As a result, an optimum task-specific leaking rate value for producing layer-specific multi-time-scale responses and a queue structure with layer-to-layer signal transmission delays for retaining past applied input enhance the Deep-ESN prediction performance. Discussion These findings can help to establish ideal design guidelines for setting the hyperparameters of Deep-ESNs.

Deep echo state network (Deep-ESN) model consists of multiple reservoir layers, that can respond to layer-specific different time-scales. This dynamical characteristic leads to enhance performance of ESN. However, neither the design guidelines for the hyperparameters of the network and individual neurons nor the mechanism to produce the diverse dynamical response have been clarified. In this study, we proposed an approach to generate the dynamical responses with different time-scales in each layer by adjusting the leaking rate of neurons. Through the evaluations time-series prediction task for different leaking rates, multiscale entropy analysis for each reservoir layer, and cross-correlation between adjacent layers, we found that when the leaking rate is set to low, the layer-specific dynamics with different time-scales are generated, as well as a mechanism whereby the signal propagates to subsequent layers with a delay, i.e., queue characteristic. These characteristics produce a high memory capacity. Consequently, diverse responses with delay leads to an enhancement of Deep-ESN functionality.

In this study, we investigate model structures for neural ODEs to improve the data efficiency in learning the dynamics of control systems. We introduce two model structures and compare them with a typical baseline structure. The first structure considers the relationship between the coordinates and velocities of the control system, while the second structure adds linearity with respect to the control term to the first structure. Both of these structures can be easily implemented without requiring additional computation. In numerical experiments, we evaluate these structure on simulated simple pendulum and CartPole systems and show that incorporating these characteristics into the model structure leads to accurate learning with a smaller amount of training data compared to the baseline structure.

We propose simple feedback control systems that work as quantum computing apparatuses applied to industrial terminal edge. The important point of quantum calculation is to keep quantum wave as superposition states. So far, the authors have proposed quantum computing by measuring the motion trajectory of macroscopic mass points. The purpose is, for example, to equip the microcomputer of the terminal of the vehicle with an unnecessary quantum computer of a large-scale device. However, it is necessary to search for a system that is easier to use than mass motion. In this paper, we show that an affine system becomes a quantum system if appropriate feedback is input, and that this is applied to quantum computation. One of the authors has developed optimal quantum mechanical feedback for nonlinear control systems. This method is applied to a time invariant linear affine control system with quadratic costs selected for concise explanation. As a result, systems that follow thermal phenomenology can be treated as quantum systems. Replace qubit and control apparatus that drives the qubit state in the current quantum computer with a fluid box and fluid temperature feedback loop. This gives a new quantum computer system without peripherals to keep, for example, cryogenic temperature.

In this paper, a support-vector-machine(SVM)- based method of detecting stenosis is presented for fallopian tubal models. It copes with stenosis detection as classification of data prepared from results of ultrasonic measurements conducted for tubal models. Under assumption that waves reflected at the second and third boundary surfaces of the models potentially include characteristics associated with blocked sections (i.e., stenosis), the method determines the time range having the reflected waves, by referring to maximal values on envelope curves of them The determined range is divided into regular short intervals, and the difference between maximum value and minimum value on envelope curves is calculated for each interval. The ten-dimensional data used to SVM learning and stenosis detection is prepared from the frequency distribution of the number of the short intervals versus difference values. Experimental results establish that the method can achieves favorable accuracies in checking occurrence of stenosis and in identifying tubal model types.

We design classical apparatuses to measure the weighting factor of eigenstates without breaking the quantum mechanical superposition state. The apparatuses work as "quantum"computer even though they follow classical physics. To do quantum computation, we need to know what percentage of a particular state is included in a superposed quantum state. However, we face difficulties that large-scale equipment, such as cryogenic temperature, is required to maintain the overlap. Combining causal formulation of quantum mechanics and mediocre classical mechanism overcomes the situation. The weighting coefficients of the superposition can be determined by monitoring the trajectory of a classical particle under quantum flucutation force. Such monitoring is possible only in macroscopic physics. Efficacy of our macroscopic apparatus model is shown in one of typical quantum algorithms, Deutsch algorithm. This classical mechanical design opens the way for "quantum mechanical"artificial intelligence to be mounted on robots in usual macroscopic world.

We design a mechanical apparatus that works as "quantum"computer even though it follows classical physics. To do quantum computation, we need to know what percentage of a particular state is included in a superposed quantum state. However, the superposition easily breaks unless expensive and large-scale tooling, such as cryogenic temperature is prepared. We pay attention to causal formulation of quantum mechanics. The weighting coefficients of the superposition can be determined by monitoring the trajectory of a quantum particle. Such monitoring is possible only in macroscopic physics. Efficacy of our macroscopic apparatus model is shown in one of typical quantum algorithms, Deutsch algorithm. This classical mechanical design opens the way for "quantum mechanical"artificial intelligence to be mounted on robots in usual macroscopic world.

In this paper, a method of achieving high accuracy in discriminating right and left eyes is proposed for ophthalmic surgery. A VGG16 convolutional neural network is employed to construct a main classifier. The data presented to the main classifier are some frames sampled at regular intervals from surgery videos. Before classifying the frames to be examined, the proposed method determines whether they are suitable or not to improve the discrimination accuracy as high as possible. In other words, the frames causing erroneous discrimination are omitted. The determination of frames depends on image characteristics associated with lightness values and edges, and on positions of eyelid speculums in them. It is based on SegNet neural networks. The proposed method determines the frames to be presented, if rectangular areas specified by bent parts of the speculums adequately appear in the predefined regions inside them. Experimental results reveal that the proposed method achieves the favorable discrimination accuracy with a small number of data in training SegNet networks compared with another method.

<p>In recent ethological studies, the behaviors and interactions of animals have been recorded by digital video cameras and webcams, which provide high functionality at reasonable cost. However, extracting the behavioral data from these videos is a laborious and time-consuming manual task. We recently proposed a novel method for tracking unmarked multiple honeybees in a flat arena, and developed a prototype software named "K-Track". The K-Track algorithm successfully resolved nearly 90% of cases involving overlapped or interacted insects, but failed when such events happened near an edge of a circular arena, which is commonly employed in experiments. In the present study, we improved our K-Track algorithm by comparing the interaction trajectories obtained from forward and backward playing of video episodes. If the tracking results differed between the forward and backward episodes, the trajectory with lower maximum moving distance per frame is chosen. Based on this concept, we developed a new software, "K-Track-kai", and compared the performances of K-Track and K-Track-kai in honeybee tracking experiments. In the cases of 6 and 16 honeybees, K-Track-kai improved the tracking accuracy from 91.7% to 96.4% and from 94.4% to 96.7%, respectively.</p>

In this paper, a method of determining examinations is proposed for ophthalmologic outpatients, using feedforward neural network (NN for short). The determination is based on the data classification. The proposed method defines four classes of ophthalmologic examinations. It prepares data for NN training and examination determination from handwriting sentences in outpatients' interview sheets. A set of the training data is prepared in the form of a matrix. The words extracted from the sentences are assigned to the matrix columns, while each sheet (or sentences in it) is assigned to a matrix rows. Entries in the matrix takes binary values meaning whether extracted words appear in the sentences in the sheet. The proposed method also the ages of outpatients as entries. NN training is conducted according to the normal backpropagation algorithm using a row as one of the training data. The trained NN has four output neurons each of which takes the value belonging to the range [0, 1]. The class of data to be examined is determined by searching the neuron at which the largest value appears in the output layer. Experimental results comprehensively establish that the proposed method can achieve higher percentages of concordance than other methods.

© 2018 IEEE. In this paper, an ultrasound-measurement-based method of detecting stenosises in fallopian tubal models is presented. The proposed method employs silicone rubber tubes and custom phantoms made of polymer gels as models of fallopian tubes. When the targets are the former (or latter), stenosis models are formed by lumps of polyester sewing threads (or polymer gels with high reflection intensity). The proposed method copes with detecting stenosis models as classification problems by applying support vector machines (SVMs for short). It prepare data for SVM learning and stenosis detection from waveforms in the range specified by two time points at which waves are probably reflected by inner upper and lower surfaces of tubal models. The data preparation is conducted by generating the frequency distribution of the number of regular short intervals into which this range is divided versus normalized values associated with amplitude of the waveforms in such intervals. When it is explored whether given data are prepared from a tube with the stenosis model, a discrimination model constructed by SVM learning works as a two-class classifier. Using the model as a multi-class classifier makes it possible to distinguish data of tubes with stenosis models from each other. It is revealed from experimental results that the proposed method has substantially high capability in detecting stenosis models.

This paper presents a cellular automaton-based model for a mechanical computer called 'Turing Tumble' computer. This computer uses mechanical reactions of a ball flowing down and mechanical components (such as gear and ramp) that are configured on the board, for its computation. A group of cells, called a supercell, is defined in the proposed model in order to represent simultaneous state transition of cells for implementing a chain reaction of connected gears. A small element with its memory called Converter is shown for an illustrative example on this model.

In this paper, a personal identification method is presented for patients visiting departments of ophthalmology. The proposed method prepares data from 32×256-sized matrices, which are inspection results measured by Optical Coherence Tomography, both at the stage of the registration and at that of the collation. It considers a three-tuple with one of top largest Pachymetry values and coordinates specifying its position in the matrix to be a characteristic point, and calculates degree of similarity between any couple of a three-tuple in the registered data and that in the collation data. It employs the sum of the above degrees to determine some subject with registered data as a person to be identified. In addition, it defines some threshold values associated with Instantaneous Keratometric, Reflective Keratometric, and Pachymetry, to eliminate false acceptance as much as possible. Experimental results establish that the proposed method can improve a false acceptance rate while keeping a high identification rate.