Search

Partial observability, or the inability of an agent to fully observe the state of its environment, exists in many real-world problem domains. However, most cognitive architectures do not have a theoretical foundation that allows for a systematic approach to handling partial observability. To address this issue, in this paper, we propose a novel cognitive architecture based on the theory of partially observable Markov decision processes (POMDPs). We focus on one of the most fundamental cognitive architectures, the blackboard architecture, and re-formulate it using the POMDP framework so that the agent can automatically decide how to use its information processing modules in partially observable environments. We propose a two-layer POMDP model for the blackboard architecture in which one layer models the dynamics of the environment and the other layer models the dynamics of the blackboard and knowledge sources within the agent. We also provide a planning method to find the optimal actions for the two-layer POMDP model based on the partially observable upper confidence bounds applied to trees (PO-UCT) algorithm. We validate the proposed architecture in both numerical and real-robot experiments.

Contribution: This article proposes a low-cost and mobile device-based learning support system for the stepper motor, which provides Web-based remote laboratory. This system allows young engineers to learn about rotational principles, control methods, and programming in stepper motors. Background: Familiarity with electrical machinery, such as motors, is an important factor for electric power engineers. Hands-on laboratories are the most effective and important learning method in the study of electrical machinery. However, the Hands-on laboratories require commitments of time, space, and significant number of expensive laboratory instruments. Intended Outcomes: In the proposed system, the students are free to repeat the experiment using a real stepper motor, for example, using a smartphone. This learning support system that combines remote laboratory and virtual laboratory is expected to be an easy-to-use for young engineers. Application Design: Proposed system consists of the low-cost microcomputers, such as raspberry Pi and Arduino, and does not use expensive software, such as LabVIEW and MATLAB. It has both low cost and compatibility with any mobile devices. Findings: The learning effectiveness and usability of the system were evaluated through questionnaires and tests for a total of 50 students at the authors’ department. Consequently, the results of the evaluation tests and questionnaires demonstrated that the proposed system will be useful to study the stepper motor.

Humanoid robots are expected to perform various jobs replacing human beings, but it is difficult for non-expert users to control them. To control humanoid robots easily, various motion imitation systems, in which a humanoid robot does the same motion as a human teacher, have been studied. In the present study, we make two contributions to building easy-to-use imitation systems. The first contribution is a flexible learning system that enables both synchronous real-time imitation, in which the robot imitates human motion almost immediately, and batch imitation, in which the robot’s motion is optimized more accurately after longer measurement of human motion. We use an efficient inference-based optimization method to enable both learning modes. The other contribution is an auditory user interface that accepts voice commands from the user and at the same time identifies the target person to be imitated. This enables the user to easily use the imitation system even when other people are present. Experimental results show that the proposed system can be used in both imitation modes and that it can identify the target person.

Recently, the optimization-by-inference approach has been proposed as a new means for solving high-dimensional optimization problems quickly. Approximate Inference COntrol (AICO) is one of the most successful and promising methods that implement the optimization-by-inference approach. AICO is able to solve stochastic optimal control problems and has already been successfully used in many applications. However, it is known that the iterative inference of AICO sometimes fails to converge to the optimal solution. To make the optimization more robust, in this paper, we propose to take model uncertainty into account. In AICO, the cost function to be minimized is accurate around a particular state of a given stochastic system, but the accuracy is uncertain in regions far from that state. Because using such an uncertain function is harmful to the convergence, we modify AICO, so that it does not use the function in uncertain regions. Our method is easy to implement and does not add much computational time to the original AICO. Experiments using two different scenarios show that our method substantially improves AICO in terms of the rate at which the algorithm produces convergent results.

One of the attractive features which a self-organizing map (SOM) possesses is a topology-preserving projection from the input layer to the competitive layer. Generally speaking, its correspondence is developed through training based on information representations of the applied multi-dimensional data. A developed feature map in the competitive layer enables us easy to understand some underlying rules visually. By the way, an analysis of Saga Prefectural sightseeing information by a SOM has been tried so far. According to the results of preceding studies, applied various topics are divided into several groups successfully. Nevertheless, there are some items not reflected in them at all. This fact implies that representations of the applied data are not appropriate to develop the feature map which we have intended in advance. Then, to overcome this tough problem, a simple idea to emphasize particular items depending on our interests is introduced as pre-synaptic processing. As a result of some computer simulations, it is confirmed that the developed feature maps are modified adaptively depending on the emphasized coefficient. And, it is concluded that the proposed simple method is effective.

The theory of partially observable Markov decision processes (POMDPs) is a useful tool for developing various intelligent agents, and learning hierarchical POMDP models is one of the key approaches for building such agents when the environments of the agents are unknown and large. To learn hierarchical models, bottom-up learning methods in which learning takes place in a layer-by-layer manner from the lowest to the highest layer are already extensively used in some research fields such as hidden Markov models and neural networks. However, little attention has been paid to bottom-up approaches for learning POMDP models. In this paper, we present a novel bottom-up learning algorithm for hierarchical POMDP models and prove that, by using this algorithm, a perfect model (i.e., a model that can perfectly predict future observations) can be learned at least in a class of deterministic POMDP environments

To improve the insulation performance of high voltage equipment, new functional materials of distributed permittivity, or the functionally graded materials (FGMs), have been developed. However, methods for measuring the permittivity distribution of these materials have yet to be established. In this study, numerical analyses have been conducted in order to investigate how the position of dielectrics in a cylindrical measurement device affects the capacitance and electric potential within the device. It has been confirmed that the capacitance and electric potential change with a varying position of the dielectrics. Therefore, it is suggested that the permittivity distribution of a given FGM can be estimated by measuring the capacitance and electric potential around it.

Calculating leakage reactance of transformers is a very important factor in designing transformers. It is difficult to calculate the leakage reactance in multiple-winding transformers, because their structures are more complicated. The finite element method (FEM) is useful to obtain the leakage reactance in such a complex case. Using a 3D FEM analysis, leakage reactances have been calculated and compared to experimental results. But, there were differences between the results of FEM analysis and the experimental results. In this paper, the leakage reactance is calculated by using a coupled-field analysis, which combined electromagnetic field analysis and electric circuit analysis, for simulation of short-circuit test in multiple-winding transformer. The results of this analysis showed that current values of parallelly connected primary windings are very different by the physical geometry of the secondary winding. The leakage reactances calculated by using the coupled-field analysis are qualitatively the same as experimental results.

We have proposed a voltage-current sensor of resin molded type for measurement of the power factor and harmonics. To understand the characteristics of the conventional prototype sensor and the problems of its practical use, we conducted a finite element analysis. Problems involving the cost and practical usage of the sensor were identified. Therefore, we have redesigned the prototype sensor, incorporating the current sensor designed on the basis of the FEM and a ready-made voltage sensor. In this paper, we describe feasibility field experiments with the present sensor in a practical 6.6-kV distribution system and show that the sensor can measure the power factor and harmonics of horizontal and vertical power line systems successfully. © 2011 Wiley Periodicals, Inc. Electr Eng Jpn, 178(3): 1–10, 2012; Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/eej.21213

We have investigated a measurement system for permittivity distribution based on the measured capacitance using a pair of long distance small size electrodes with the shield and arc additional electrodes. However, the estimation of the permittivity distribution has not been established yet. In the paper, we carried out experiments and numerical analyses in order to verify the effect of size and position of the arc additional electrode and discussed the effect of measured capacitance on the arc additional electrode.

EGFR mutations are a major determinant of lung tumor response to gefitinib, an EGFR-specific tyrosine kinase inhibitor. Obtaining a response from lung tumors expressing wild-type EGFR is a major obstacle. The combination of gefitinib and cytotoxic drugs is one strategy against lung cancers expressing wild-type EGFR. The DNA topoisomerase inhibitor irinotecan sulfate (CPT-11) is active against lung cancer. We examined the sensitivity of lung cancers expressing wild- or mutant-type EGFR to the combination of gefitinib and CPT-11. The in vitro effect of gefitinib and SN-38 (the active metabolite of CPT-11) was examined in seven lung cancer cell lines using the dye formation assay with a combination index. When administered concurrently, gefitinib and SN-38 had a synergistic effect in five of the seven cell lines expressing wild-type EGFR, whereas the combination was antagonistic in PC-9 cells and a PC-9 subline resistant to gefitinib and expressing deletional mutant EGFR (PC-9/ZD). When administered sequentially, treatment with SN-38 followed by gefitinib had remarkable synergistic effects in the PC-9 and PC-9/ZD cells. In an in vivo tumor-bearing model, this combination had a schedule-dependent synergistic effect in the PC-9 and PC-9/ZD cells. An immunohistochemical analysis of the tumors in mice treated with CPT-11 and gefitinib demonstrated that the number of Ki-67 positive tumor cells induced by CPT-11 treatment was decreased when CPT-11 was administered in combination with gefitinib. In conclusion, the sequential combination of CPT-11 and gefitinib is considered to be active against lung cancer.

The number of published studies on peritoneal dissemination of scirrhous gastric carcinoma is very small as a result of the unavailability of highly reproducible animal models. Orthotopic implantation of HSC-44PE and HSC-58 (scirrhous gastric carcinoma-derived cell lines) cells into nude mice led to dissemination of the tumor cells to the greater omentum, mesenterium, peritoneum and so on, and caused ascites in a small number of animals. Cycles of isolation of the ascitic tumor cells and orthotopic inoculation of these cells were repeated in turn to animals. This was to isolate highly metastatic cell lines with a strong capability of inducing the formation of ascites (44As3 from HSC-44PE; 58As1 and 58As9 from HSC-58). All three cell lines induced tumor formation at the site of orthotopic injection, and caused fatal cancerous peritonitis and bloody ascites in 90-100% of the animals approximately 3-5 weeks after the inoculation. When the parent cells were implanted, the animals became moribund in approximately 12-18 weeks, however, none of the animals developed ascites. Complementary DNA microarray and immunohistochemical analyses revealed differences in the expression levels of genes coding for the matrix proteinase, cell adhesion, motility, angiogenesis and proliferation between the highly metastatic- and parent-cell lines. The usefulness of this model for the evaluation of drugs was assessed by analyzing the stability of the metastatic potential of the cells and the reproducibility. Animals intravenously treated with CPT-11 and GEM showed suppressed tumor growth and significantly prolonged survival. The metastatic cell lines and the in vivo model established in the present study are expected to serve as a model of cancerous peritonitis developing from primary lesions, and as a useful means of clarifying the pathophysiology of peritoneal dissemination of scirrhous gastric carcinoma and the development of drugs for its treatment.

Vinorelbine (Navelbine, KW-2307), a semisynthetic vinca alkaloid, is a potent inhibitor of mitotic microtubule polymerization. The aims of this study were to demonstrate vinorelbine-induced radiosensitization of human small cell lung cancer (SCLC) SBC-3 cells and to elucidate the mechanisms of radiosensitization. A clonogenic assay demonstrated that SBC-3 cells were sensitized to radiation by vinorelbine using different schedules combining exposure to both. The sensitizer enhancement ratios (SERs) at a cell survival level of 10% were 1.42+/-0.21 to 1.33+/-0.06, and 1.22+/-0.07 depending on schedule. Vinorelbine-induced radiosensitization did not depend on the schedule of the combined exposure. Flow cytometric analyses showed that the cells did not accumulate in the radiosensitive G(2)/M phase of the cell cycle after concurrent treatment with vinorelbine and radiation. The results of an alkaline filter elution assay demonstrated that in the presence of vinorelbine at 1 n M radiation-induced DNA strand breaks were not completely repaired at 24 h postradiation. We conclude that human SCLC SBC-3 cells are sensitized to radiation by vinorelbine and that a possible mechanisms of vinorelbine-induced radiosensitization may at least in part be associated with impairment of DNA repair following radiation-induced DNA damage.

Angiogenesis is an essential process for tumor growth and is regulated by tumor-derived angiogenic cytokines. Osteopontin (OPN) is one of the cytokines produced by various tumor cells and is suggested to be involved in angiogenesis by upregulating endothelial cell migration in cooperation with vascular endothelial cell growth factor (VEGF). To provide evidence of OPN involvement in a causal role in tumor angiogenesis, we generated a stable transfectant from murine neuroblastoma C1300 cells to constitutively secrete high levels of murine OPN. The OPN mRNA expression and protein secretion were confirmed by RT-PCR and ELISA, respectively. The biological activity of secreted OPN was determined with migration assay by using human umbilical vein endothelial cells (HUVEC). Transfection with OPN gene did not increase VEGF production and did not affect gene expression of other angiogenic factors confirmed by complementary DNA macroarray system. To demonstrate the effect of OPN on tumor-induced angiogenesis in vivo, millipore chambers containing OPN-transfected or control cells were implanted to the dorsal air sac of mice. The OPN-transfected cells significantly induced neovascularization in comparison to the control cells in mice. Conclusively, these results provide direct evidence of OPN involvement in the role of tumor angiogenesis. © 2002 Wiley-Liss, Inc.

Gangliosides such as GD3, GM2, and GD2 are abundantly expressed on the cell surfaces of various malignant cells, suggesting the potential for anti-ganglioside antibody therapy for tumors. Anti-ganglioside GD2 antibody treatment is currently undergoing clinical trials for melanoma and neuroblastoma. We previously reported high in vivo antitumor effects of anti-GM2 ganglioside antibody against lung cancer. To determine whether anti-GM2 antibody may be clinically indicated for gastrointestinal cancers, we evaluated the mRNA expression of α2,8 sialyltransferase, a GD3 synthase, and β1,4 N-acetylgalactosaminyltransferase (β1,4 GalNAc-T), a GM2/GD2 synthase, in gastrointestinal cancers. We performed modified semi-quantitative RT-PCR, which reduces complexity incidental to radiolabeling on samples taken from small surgically removed clinical specimens. Stomach (19/22) and colorectal (21/30) cancers showed decreased expression of α2,8 sialyltransferase as compared with respective normal tissues (P < 0.05). In contrast, increased expression of β1,4 GalNAc-T was detected in both types of tumors. Clinicopathological analysis revealed significantly higher expression level of α2,8 sialyltransferase in the poorly differentiated than in the well-differentiated stomach cancer group (P < 0.05). Furthermore, the expression level of α2,8 sialyltransferase was significantly decreased in male as compared with female colorectal cancer patients (P < 0.05). These results suggest that expression level of GM2 ganglioside is elevated in gastrointestinal cancer, and that anti-GM2 antibody may be applicable to its treatment.

The oncogenenic transmembrane tyrosine kinase receptor HER-2/neu is a promising target for treatment of HER-2-overexpressing cancers. The humanized anti-HER-2/neu antibody Trastuzumab is under clinical evaluation in combination with chemotherapy against breast cancer. The combination of Trastuzumab and cisplatin is expected to be active against HER-2 / neu-expressing tumors. We examined the mechanisms of this combination effect against human solid tumor cells in the presence of human peripheral blood mononuclear cells (PBMCs) using an in vitro MTT assay. The growth-inhibitory effects of cisplatin (CDDP) on the tumor cells were not significantly affected by Trastuzumab in the absence of effector cells. CDDP alone at a dose of less than 12.5 mM did not affect the viability of PBMCs, as determined by MTT assay, suggesting that PBMCs could exert antibody-dependent cell-mediated cytotoxicity (ADCC) at this CDDP concentration. The combination of Trastuzumab and CDDP showed higher cytotoxic effects against the tumor cells in the presence of PBMCs. The CDDP concentration required to inhibit tumor cell growth by 50% was reduced to approximately 20% by Trastuzumab in the presence of PBMCs at an effector/target ratio of 10. It may be important to select combined chemotherapeutic agents which do not diminish the ADCC activity of Trastuzumab via PBMCs. Both the expression of HER-2 / neu and the ADCC activity may be important determinants of the therapeutic benefit of the Trastuzumab/CDDP combination.

Purpose: TZT-1027, an antimicrotubule agent that inhibits the polymerization of tubulin, shows potent antitumor activity in various transplantable tumor models in vivo. The high antitumor activity of TZT-1027 prompted us to speculate that this compound may have a mode of action other than its antimicrotubule and antimitotic activities. To elucidate the interaction of antitumor cytokines with TZT-1027 in tumors in vivo, we examined the antitumor activity of this agent against various cytokine gene-transfected Lewis lung carcinoma (LLC) cells inoculated into C57BL/6 mice. Methods: In vitro growth inhibition was evaluated using the MTT assay, and in vivo activity was evaluated in subcutaneous models in C57BL/6 mice. The status of the vasculature in tumor tissues was evaluated immunohistochemically using anti-CD31 antibody. We used a cDNA macroarray to examine the gene expression profiles in tumor tissues removed from mice. Results: TZT-1027 at 3 mg/kg showed potent antitumor activity in Mock (LLC-Neo cells) inoculated mice with a T/C% value of 16%. TZT-1027 at 3 mg/kg showed more potent antitumor activity in LLC-TNF cells and LLC-IL6 cells with T/C% values of 4% and 3%, respectively. TZT-1027 treatment destroyed the tumor vasculature as well as tumor cells in LLC-TNF and LLC-IL6 tissues of mice treated with TZT-1027. The LLC-TNF and LLC-IL6 tissues of mice treated with TZT-1027 had in common the independent alteration of the non-histone chromosomal protein HMG-14 and transcription factor 1 for heat shock gene. Focusing on the gene regulation related to angiogenesis, the alteration in transcriptional factors such as ets family genes and homeobox family genes was remarkable. Conclusions: These factors are candidates as determinants of the enhanced TZT-1027 antitumor activity in relation to these cytokines.

Calculating leakage reactance of transformers is a very important factor in designing transformers. In the traditional two-winding transformer, the calculation methods of leakage reactance have been presented using the inductance deduced by the geometry mean length between two windings. However, it is difficult to calculate the leakage reactance in multi-winding transformers, because its structure is more complex. Finite element method (FEM) is useful to obtain the characterization of the electromagnetic behavior of the magnetic components. FEM is an accurate method but its computational cost is very high. In this paper, the leakage reactance of 9-winding transformer was calculated by using an ordinary three dimensional FEM (3D FEM) and hybrid FEM which is a quasi-three dimensional analysis method. The comparison results show that the hybrid FEM is able to calculate the leakage reactance with a low computational cost.

The current sensor of the voltage-current sensor system proposed by us is two air-core sensing coils connected in series to each other with an opposite polarity. By doing so, it can measure the harmonics and power factor while removing the influence from power lines of other phases. When the sensors are surrounded by an iron case, EMF may be influenced. However, it has been confirmed that a large change is not observed with regard to measurement of harmonics, and the proposed sensors can be used without problem. Therefore it was found out that the proposed sensor is able to perform power measurements in a state of being incorporated in the pole switch.

In recent years, “Smart Grid” obtains much attention. In the our laboratory, “Simulated Micro Grid System” for supporting learning of power engineering has been designed and implemented. The system helps young engineers to understand the principles of the Smart Grid technology. In this study, we have developed a power measurement system for our Simulated Micro Grid System. AD conversion is essential in the power measurement system to control components in the Simulated Micro Grid System and deliver measured data on the Web. In this paper, we have proposed two methods, AD conversion using ART-Linux and AD conversion using SH-3 microcomputer. The method by ART-Linux is used when real-time control is required. Moreover, the method by SH-3 microcomputer is used when there are many measurement targets.

Experiments have critical roles in engineering education. However, it is not easy to conduct experiments because of their high costs of equipment and unavailability. Remote experiment systems have coped with this issue. Using remote experiment systems, users can conduct experiments regardless of their locations or times. Remote experiment systems have been developed worldwide, but most of the systems are developed for personal computers, and therefore cannot be used with smartphones or tablets. This study aims to develop a remote experiment system which can be used with smartphones or tablets. The proposed system provides remote control of a DC motor by PWM control and remote measurement of a single phase alternator. The user interface of the system is designed without Web plug-ins so that users can use the systems with their smartphones or tablets as well as their personal computers.

Selective tyrosine kinase inhibitors are regarded as promising antitumor agents for cancer treatment. Iressa (ZD1839) is an orally active, selective EGFR-TKI (epidermal growth factor receptor-tyrosine kinase inhibitor) that blocks signal transduction pathways implicated in cancer cell proliferation, survival and other host-dependent processes promoting cancer growth. The cellular mechanisms of ZD1839 action against human malignant cells and drug-resistant cells were evaluated in vitro. Among the cell lines tested, ZD1839 showed a strong growth-inhibitory effect in vitro on human leukemic cells resistant to phorbol ester. This cell line, K562/TPA, shows a non-P-glycoprotein-mediated multidrug-resistant phenotype. The IC50 value of ZD1839 on K562/TPA was approximately 400-fold lower than that on the parental K562 cell (K562 = 12 +/- 2 microM; K562/TPA = 0.025 +/- 0.002 microM) in vitro as determined by a dye formation assay. The expression of EGFR and EGFR mRNA was clearly present in K562/TPA but not in parental K562 cells as determined by Western blotting and RT-PCR. EGFR was autophosphorylated in K562/TPA detected by the antiphosphotyrosine antibody. The in vivo antitumor effects of ZD1839 on K562 and K562/TPA cells were also investigated in BALB/c nude mice. K562/TPA cells transplanted subcutaneously into mice disappeared completely with ZD1839 treatment (20 mg/kg/day, days 3-9). This was not the case in K562 cells. These results suggest that ZD1839 is highly active against tumor cells with non-P-glycoprotein-mediated multidrug resistance that express EGFR. Iressa is a trademark of AstraZeneca (Cheshire, UK).