A brainlike learning system with supervised,unsupervised, and reinforcement learning

According to Hebb's cell assembly theory, the brain has the capability of function localization. On the other hand, it is suggested that in the brain there are three different learning paradigms: supervised, unsupervised, and reinforcement learning, which are related deeply to the three parts of brain: cerebellum, cerebral cortex, and basal ganglia, respectively. Inspired by the above knowledge of the brain in this paper we present a brainlike learning system consisting of three parts: supervised learning (SL) part, unsupervised learning (UL) part, and reinforcement learning (RL) part. The SL part is a main part learning input–output mapping; the UL part is a competitive network dividing input space into subspaces and realizes the capability of function localization by controlling firing strength of neurons in the SL part based on input patterns; the RL part is a reinforcement learning scheme, which optimizes system performance by adjusting the parameters in the UL part. Numerical simulations have been carried out and the simulation results confirm the effectiveness of the proposed brainlike learning system. © 2007 Wiley Periodicals, Inc. Electr Eng Jpn, 162(1): 32–39, 2008; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/eej.20600     

A novel method for induction and detection of anaphylactic reaction using the mouse abdominal wall (AW method)

We found that an antigen-specific anaphylaxis was induced by antigen challenge to the abdominal wall, ear auricle, or subcutaneous tissue in mice sensitized 9 days previously with antigen and adjuvant. The anaphylactic reaction was detected by vascular permeability at the injected site 7 minutes after challenge, which was the best time for estimation. A novel method (AW method) for induction and detection of the anaphylactic reaction in mice was established using the abdominal wall as the challenge site. This method could detect the anaphylactic response in mice 1 to 3 weeks after sensitization. The increase in vascular permeability was completely inhibited by administration of diphenhydramine.     

Character motion synthesis by principal component analysis and motion control interface by hands

In this paper, we propose an interactive character motion control interface that uses hands. Using their hands and fingers, the user can control a large number of degrees of freedom at the same time. We applied principal component analysis to a set of sample poses and assigned the extracted principal components to each degree of freedom of the hands (such as the hand positions and finger bending/extending angles). The user can control the blending weights of the principal components and deform the character's pose by moving their hands and bending/extending their fingers. We introduced pose and action controls, so that we can alter the standing pose and perform various actions with deformations. So that various types of actions were possible, we constructed a number of action models in advance. We introduced action model selection and action execution mechanisms. We developed methods for computing the feature vector, for applying principal component analysis, and for pose and action synthesis. In addition, we introduced a pose transition method for performing a step motion when necessary to prevent foot sliding. We present our experimental results and demonstrate the effectiveness of our interface. Copyright © 2015 John Wiley & Sons, Ltd.     

CSD-controlled reactive crystallization of SrSO4 in the presence of polyethylenimine

Reactive crystallization of SrSO₄ was performed in the presence of polyethylenimine (PEI) using a double-jet crystallizer. PEI dosage suggested important dependence on crystal size distributions (CSDs) leading to obtain micron-sized crystals with mono-dispersed distribution. A plausible role of PEI was considered to be a multiple inhibitor of growth, continuous nucleation and agglomeration. PEI might be an effective polyelectrolyte additive for the production of mono-dispersed fine crystals.     

Preparation of mesoporous silica replica using ordered mesoporous carbon by vapor phase transport of silica source

A new vapor phase transport (VPT) technique to prepare an inverse silica replica of an ordered mesoporous carbon was developed. Tetraethyl orthosilicate (TEOS) was infiltrated in mesoporous carbon CMK-3 as the hard template at 180 °C for 48 h under an autogenous pressure in an autoclave in the presence of water. The samples obtained by removal of CMK-3 retained structural regularity of CMK-3 with little shrinkage of framework, which were characterized by SAXRD, N₂ adsorption, TG-DTA, and SEM. Influence of preparation temperature on the loading amount of silica was investigated. The multi-step replication process was monitored by characterizing the replicated materials as well as intermediate composites.     

Molded micro- and mesoporous carbon/silica composite from rice husk and beet sugar

A molded carbon/silica composite with high micro- and mesoporosity, as well as a high bulk density, was fabricated by activating a disk-molded precursor made from carbonized rice husk (RH) and beet sugar (BS) at 875 °C in CO₂. The pore structure of the RH- and BS-based carbon/silica composite (RBC) was analysed in relation to the bulk density. An activation time of 2.0 h provided the largest BET specific surface area (1027 m²/g) and total pore volume (0.68 cm³/g) and a low bulk density (0.54 g/cm³). An RBC that was first activated for 1 h was immersed again in BS syrup and then activated in CO₂ for 1 h. This two-step activation process provided both a high bulk density (0.69 g/cm³) and a highly textured structure (BET specific surface area, 943 m²/g; total pore volume, 0.56 cm³/g). The immersion in BS syrup was useful for improving the texture without reducing the bulk density, in comparison to one-step activation for 1.0 h. The suspension of the RBCs was basic because of the residual inorganic compounds of potassium and calcium. However, the basicity of the suspension was alleviated by washing the RBCs with water.     

Disorders and oxygen vacancies in p-type Sn2B2O7 (B = Nb,Ta): Role of the B-site element

Sn₂Nb_2−xTa_xO₇ (x = 0.0–2.0) with pyrochlore structure is a promising material for p-type oxide semiconductors. A systematic study of its Nb/Ta ratio indicated that the hole–generation efficiency of the Nb end (Sn₂Nb₂O₇) was an order of magnitude lower than that of the Ta end (Sn₂Ta₂O₇). Although this occurs due to differences in oxygen-vacancy formation, the origins of the hole–generation efficiencies remain unclear due to limited information on local and global crystal-structure disorders in pyrochlore Sn₂Nb₂O₇ and Sn₂Ta₂O₇. In this study, the crystal structures of Sn₂B₂O₇ (B = Nb, Ta), composed of BO₆ octahedra and Sn₄O tetrahedra, were investigated using X-ray absorption spectroscopy and X-ray diffraction. A detailed investigation of the local and global crystal structures indicated a larger amount of disorder in the Sn₄O tetrahedra in Sn₂Nb₂O₇ compared to Sn₂Ta₂O₇; disorder in the BO₆ octahedra occurred only in Sn₂Ta₂O₇. This study indicates that an appropriate selection of the B-site element is vital for suppressing defect and disorder formation in Sn₄O tetrahedra and subsequently improving the hole–carrier–generation efficiency.     

Improvement of the surface quality of foam injection molded products from a material property perspective

Microcellular injection molding is an attractive method. However, their surface imperfections have been a major problem hindering wide industrial applications. Several methods have been proposed to improve the surface appearance of foams. In this study, we proposed a method to improve the surface appearance of polypropylene (PP) foams from the material property perspective, especially with regard to crystallization and viscosity. The basic idea of the surface improvement is to reduce the size of bubbles generated at the flow front, delay the solidification behavior of the polymer at the mold interface, squeeze the bubbles existing at the mold–polymer interface, and redissolve the bubbles into the polymer by holding pressure. Blending a low-modulus PP delays the crystallization of the polymers at the skin layer and solidification, taking enough time to squeeze the bubbles smaller. A sorbitol-based gelling agent, bis-O-([4 methylphenyl]methylene)-D-Glucitol, was used to increase the viscosity at a low strain rate to reduce the size of the bubbles generated at the flow front during the filling stage. The foam injection molding experiments demonstrated that the proposed method effectively improved the surface appearance of the foams. In particular, the surface appearance of the foams became almost equivalent to that of solid samples using low-modulus PP.     

Acquisition of optimal attribute subset through genetic algorithm using GNP‐based class association rule mining

Attribute selection is a technique to prune less relevant information and discover high‐quality knowledge. It is especially useful for the classification of a large database, because the preprocessing of data increases the possibility that predictor attributes given to the mining algorithm become more relevant to the class attribute. In this paper, a method to acquire the optimal attribute subset for the genetic network programming (GNP) based class association rule mining has been proposed, and this attribute selection process using genetic algorithm (GA) leads to a higher accuracy for classification. Class association rule mining through GNP is conducted with a small subset of data rather than the original large number of attributes; thus simple but important rules are obtained for classification while the local optimal problem is avoided. Simulation results with educational data show that the classification accuracy is largely improved from 52.73 to 74.54%, when classification is made using the optimal attribute subset. © 2014 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.