A dynamic associative memory system by adopting an amygdala model

Although several kinds of computational associative memory models and emotion models have been proposed since the last century, the interaction between memory and emotion is almost always neglected in these conventional models. This study constructs a dynamic memory system, named the amygdala-hippocampus model, which intends to realize dynamic auto-association and the mutual association of time-series patterns more naturally by adopting an emotional factor, i.e., the functional model of the amygdala given by Morén and Balkenius. The output of the amygdala is designed to control the recollection state of multiple chaotic neural networks (MCNN) in CA3 of the hippocampus-neocortex model proposed in our early work. The efficiency of the proposed association system is verified by computer simulation using several benchmark time-series patterns. This work was presented in part at the 13th International Symposium on Artificial Life and Robotics, Oita, Japan, January 31–February 2, 2008     

Intelligent agent construction using the attentive characteristic patterns of chaotic neural networks

Humans learn from incidents in their own life and reflects these in subsequent actions as their own experiences. These experiences are memorized in the brain and recollected when necessary. This research incorporates this type of intelligent information processing mechanism and applies it to an autonomous agent. In the proposed system, the reinforcement Q-learning method is used. Autoassociative chaotic neural networks are also used as mutual associative memory systems. However, an agent cannot retrieve all stored patterns exactly, especially in the case of too many stored patterns and a strong correlation among them. To solve this problem, we propose to use types of attentive parameters and attentive characteristic patterns. The attentive characteristic pattern is part of the stored patterns. When robots concentrate their attention on a specific part of a stored pattern, i.e., the attentive characteristic pattern, whole stored patterns are retrieved easily and completely. Finally, the effectiveness of the proposed method is verified through a simulation applied to plural maze-searching problems.     

Continuous-Flow Asymmetric Hydrogenation for the Synthesis of an (S)-Metolachlor Intermediate Using Catalysts Immobilized on a Core/Shell-Type Support

Here we show continuous-flow synthesis of a key inter-mediate of (S)-Metolachlor, which is a widely used agrochemical that is in huge demand, with asymmetric hydrogenation using a heterogeneous catalyst immobilized on a core/shell-type support. We have immobilized a homogeneous catalyst onto a core/shell-type heterogeneous support that was developed in our laboratory, which showed high turnover frequency and turnover number over a long period (over 120 h) when used in a solvent-free continuous-flow manner. This result demonstrates the feasibility of continuous-flow production of chiral agrochemicals with packed-bed heterogeneous catalysts. Moreover, the heterogenization protocol was shown to be applicable to elaborate practical catalysts used in industry, especially the group of phosphine ligands.     

A robust reinforcement learning using the concept of sliding mode control

In this article, we propose a new control method using reinforcement learning (RL) with the concept of sliding mode control (SMC). Some remarkable characteristics of the SMC method are good robustness and stability for deviations from control conditions. On the other hand, RL may be applicable to complex systems that are difficult to model. However, applying reinforcement learning to a real system has a serious problem, i.e., many trials are required for learning. We intend to develop a new control method with good characteristics for both these methods. To realize it, we employ the actor-critic method, a kind of RL, to unite with the SMC. We are able to verify the effectiveness of the proposed control method through a computer simulation of inverted pendulum control without the use of inverted pendulum dynamics. In particular, it is shown that the proposed method enables the RL to learn in fewer trials than the reinforcement learning method. This work was presented in part at the 13th International Symposium on Artificial Life and Robotics, Oita, Japan, January 31–February 2, 2008     

Handwriting character classification using Freeman’s olfactory KIII model

KIII model is an olfactory model proposed by W. J. Freeman referring to a physiological structure of mammal??s olfactory system. The KIII model has been applied to kinds of pattern recognition systems, for example, electronic nose, tea classification, etc. However, the dynamics of neurons in the KIII model is given by Hodgkin-Huxley??s second-order differential equation and it consumes a very high computation cost. In this paper, we propose a simplified dynamics of chaotic neuron instead of the Hodgkin-Huxley dynamics at first, and secondly, we propose to use Fourier transformation with high resolution capability to extract features of time series behaviors of internal states of M1 nodes in KIII model instead of the conventional standard deviation method. Furthermore, paying attention to the point that human brain does visual processing as same as olfactory processing in the sense of information processing, a handwriting image recognition problem is treated as a new application field of KIII model. Through the computer simulation of the handwriting character classification, it is shown that the proposed method is useful by the comparison of experiment results with both computation time and recognition accuracy.     

Clinical electroencephalographic study of nine pediatric patients with convulsion induced by the TV animation, Pocket Monster

Unruptured abdominal aortic aneurysm (AAA) is seldom recognized. Thus it is difficult to know whether the incidence of AAA in the general population is high enough to warrant routine screening at least in men after a certain age. Ultrasound screening studies to evaluate the incidence of AAA have been carried out in several English-speaking and Scandinavian countries. The purpose of this report is to describe the results of a study carried out in Belgium. All 65- and 75-year-old men living in the city of Liege, Belgium, were given the opportunity to undergo a free ultrasound examination. Only 41% of the target population was examined. AAA defined as abdominal aortic diameter of >30 mm was observed in 28 subjects (incidence: 3.8%). Mean abdominal aortic diameter was 34.7 mm. A diameter >29 mm was observed in 33 subjects (incidence 4.5%). Mean abdominal aortic diameter was 30.4 mm. On the basis of epidemiological data collected, a high-risk population for AAA was identified. Arterial hypertension (p < 0.05), previous coronary artery surgery (p < 0.05), and smoking (p < 0.06) were more common in subjects with than without AAA. The overall cost of screening was $18.175. The cost per AAA diagnosed was $551.00.     

A transient chaotic associative memory model with temporary stay function

When chaotic dynamics is imparted to the neurons that compose the associative memory model, they search for stored patterns in a pattern space chaotically. However, this model has the deficiency that judgment of whether the stored pattern has been recollected or not is difficult because its behavior is always chaotic. Because all dynamics of the chaotic neurons are chaotic, chaotic transition is repeated. The transient‐chaotic associative network (TCAN) that Lee proposed changes from the state of chaos to the state of stability (nonchaos) transiently. Additionally, it has fast recollection speed, and has large memory capacity. However, the states of TCAN do not change chaotically. Based on these results, this paper proposes a transient chaotic associative memory model with a temporary stay function (TCAMMwithTSF) which has two capabilities: one is fast speed as the states of the model converge to a stored pattern, like TCAN, and the other is the ability to search the stored pattern in a pattern space chaotically, like chaotic neural networks. Finally, the characteristics and usefulness of TCAMMwithTSF are verified by a simulation study. © 2011 Wiley Periodicals, Inc. Electr Eng Jpn, 175(2): 29–36, 2011; Published online in Wiley Online Library ( wileyonlinelibrary.com ). DOI 10.1002/eej.21077     

Investigating the Effects of Transfer Learning on ROI-based Classification of Chest CT Images: A Case Study on Diffuse Lung Diseases

Research on Computer-Aided Diagnosis (CAD) of medical images has been actively conducted to support decisions of radiologists. Since deep learning has shown distinguished abilities in classification, detection, segmentation, etc. in various problems, many studies on CAD have been using deep learning. One of the reasons behind the success of deep learning is the availability of large application-specific annotated datasets. However, it is quite tough work for radiologists to annotate hundreds or thousands of medical images for deep learning, and thus it is difficult to obtain large scale annotated datasets for various organs and diseases. Therefore, many techniques that effectively train deep neural networks have been proposed, and one of the techniques is transfer learning. This paper focuses on transfer learning and especially conducts a case study on ROI-based opacity classification of diffuse lung diseases in chest CT images. The aim of this paper is to clarify what characteristics of the datasets for pre-training and what kinds of structures of deep neural networks for fine-tuning contribute to enhance the effectiveness of transfer learning. In addition, the numbers of training data are set at various values and the effectiveness of transfer learning is evaluated. In the experiments, nine conditions of transfer learning and a method without transfer learning are compared to analyze the appropriate conditions. From the experimental results, it is clarified that the pre-training dataset with more (various) classes and the compact structure for fine-tuning show the best accuracy in this work.

     

A learning Fuzzy Petri net model

Fuzzy Petri net (FPN) is a powerful modeling tool for the construction of knowledge systems. In this paper, we propose a new learning model tool—learning fuzzy Petri net (LFPN). In contrast with the existing FPN, there are three extensions in the new model: (i) the place can possess different tokens which represent different propositions; (ii) these propositions have different degrees of truth toward different transitions; and (iii) the truth degree of the proposition can be learned by adjusting the arc's weight function. The LFPN model obtains the capability of learning fuzzy production rules through truth degree updating. The LFPN learning algorithm which introduces network learning method into Petri net update is proposed and the convergence of algorithm is analyzed. Finally, for the purpose of certification of the effectiveness of the proposed model, LFPN is used to model the Web service discovery. The result of the simulation shows that the proposed LFPN is useful and effective. © 2012 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.     

A method for analyzing the spatiotemporal changes of chaotic neural networks

A chaotic neural network proposed (CNN) by Aihara et al. is able to recollect stored patterns dynamically. But there are difficult cases such as its long time processing of association, and difficult to recall a specific stored pattern during the dynamical associations. We have proposed to find the optimal parameters using meta-heuristics methods to improve association performance, for example, the shorter recalling time and higher recollection rates of stored patterns in our previous works. However, the relationship between the different values of parameters of chaotic neurons and the association performance of CNN was not investigated clearly. In this paper, we propose a method to analyze the spatiotemporal changes of internal states in CNN and, by the method, analyze how the change of values of internal parameters of chaotic neurons affects the characteristics of chaotic neurons when multiple patterns are stored in the CNN. Quantile–Quantile plot, least square approximation, hierarchical clustering, and Hilbert transform are used to investigate the similarity of internal states of chaotic neurons, and to classify the neurons. Simulation results showed that how different values of an internal parameter yielded different behaviors of chaotic neurons and it suggests the optimal parameter which generates higher association performance may concern with the stored patterns of the CNN.