-
Kazuhiro MATSUO
Article type: Preface
1998 Volume 13 Issue 4 Pages
505
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Naohiko URAMOTO
Article type: Cover article
1998 Volume 13 Issue 4 Pages
506
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Makoto MURATA
Article type: Special issue
1998 Volume 13 Issue 4 Pages
507-514
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Kent TAMURA
Article type: Special issue
1998 Volume 13 Issue 4 Pages
515-518
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Naohiko URAMOTO, Koichi TAKEDA
Article type: Special issue
1998 Volume 13 Issue 4 Pages
519-527
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Koiti HASHIDA
Article type: Corner article
1998 Volume 13 Issue 4 Pages
528-535
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Katsuhiko TSUJINO, Tohru KOYAMA, Mie NAKATANI
Article type: Corner article
1998 Volume 13 Issue 4 Pages
536-545
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Tetsuo ONO, Satoshi TOJO
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
546-559
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
We propose a model of common language acquisition, assuming an artificial community that consists of multiple agents ; each agent owns different grammar rules at the first stage, however they come to organize a common grammar in the community through a number of exchanges of sentences between them. We hypothesize that each agent has ability of abductive and inductive inference. In the early stage of the life-span of agents, they try to generate grammar rules abductively, to parse other's sentences, however in the later stage, they try to find rules inductively from a number of sentence examples they acquired so far. The communicative ability of agents is measured by energy. The energy score of each agent becomes high if he/she could recognize others' sentences, or his/her utterances could be recognized by others. According to this energy score, each agent changes his/her behavior ; when the score is high, he/she can increase the chances of utterance, anb can give more influence upon the grammar of the whole community. As each agent modifies his/her grammar by the inferences, the common grammar in the community keeps changing dynamically. In our computer simulation, we could show (1) that the adaptability and the robustness of the common grammar were realized if abductive/inductige inferences were adequately combined, (2) that the common grammar was developed when agents selected the proper inference autonomously, and (3) that the fusion and the bifurcation of grammar emerged through a larger-scale, longer-term experiment.
View full abstract
-
Kazuhiro CHIBA, Kunihiko HIRAISHI
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
560-569
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
The purpose of this paper is to propose a new formalization of the prisoner's dilemma game, called the iterated continuous prisoner's dilemma game, and to show its usefulness. In recent years, the prisoner's dilemma game has been used in many researches which aim to investigate the principles of cooperation among agents. In the prisoner's dilemma game, each player chooses his action from two alternatives, Cooperate and Defect. However, this is unsatisfactory if we consider various kinds of human interactions, i. e., there are cases in which each player shows an inexact attitude. Therefore, we attempt to extend the range from which a player takes his action so as to represent such situations. First, we give a formulation of the continuous prisoner's dilemma game. It is a two-person non-zero-sum game in which each player can choose his action from the range [0, 1], where the value closer to 0 means more cooperative and the value closer to 1 means the player takes an action closer to defect. In the usual prisoner's dilemma game a payoff of each player is defined by a matorix, while it is given by payoff functions in the continuous version. The game is played repeatedly. This is called the iterated continuous prisoner's dilemma game. We can discuss usefulness of this new formalization by investigating various strategies each player can take on this game. Each strategy is evaluated by a present value of a total payoff under a given discount rate. Using this value, we can define a situation of invasion between two strategies. We try to compare various strategies by this invasion relation. Tit for Tat (TFT) is a well-known good strategy in the iterated prisoner's dilemma game. We evaluate various strategies against TFT, discrete TFT, and strategies that take inexact values. Lastly, we conclude this paper by discussing the principles of interaction between agents.
View full abstract
-
Koichi TAKAHASHI, Hitoshi SENO
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
570-579
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
The fractal condensation of images provides informations of fractal structures or self similarities in their contractive Affin transformation coefficients. Statistical analyses of the fractal informations are used to discriminate the image features of cancer cells. We discuss the possibility to build the diagnosis system of cell morphologies with the present method.
View full abstract
-
Akio GOFUKU, Hiroyuki HATADA, Yutaka TANAKA
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
580-587
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
An intelligent numerical simulation called self-adaptive numerical simulation (SANS) is proposed to execute efficient numerical calculations in conceptual designs or safety analyses of engineering systems. The key features of the SANS are that i) suitable model fragments are selected from a model fragments' knowledge-base, ii) a simulation program is automatically generated from the selected model fragments, iii) the execution of a simulation is postponed when the calculation results show that the conditions and applicability range of model fragments are violated, and iv) the simulation is restarted by automatically regenerating the simulation program suitable to a new situation. By these features, one can execute a numerical simulation using suitable mathematical models throughout a whole simulation time. This paper describes a framework to realize the SANS, a knowledge-base of model fragments, three key mechanisms, and a prototype system implemented only the core techniques. The SANS is demonstrated by a simple example of heating the water flow in a pipe.
View full abstract
-
Jun MIURA, Yoshiaki SHIRAI
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
588-596
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
This paper proposes a planning method for a vision-guided mobile robot under limited computational resources and vision uncertainty. The method considers the following two trade-offs : (1) granularity in approximating a probabilistic distribution vs. plan quality, and (2) search depth vs. plan quality. The first trade-off is managed by predicting the plan quality for a granularity using a learned relationship between them, and by adaptively selecting the best granularity. The second trade-off is managed by formulating the planning process as a search in the space of feasible plans, and by appropriately limiting the search considering the merit of each step of the search. Simulation results show the feasibility of the method.
View full abstract
-
Kazuhisa SETA, Mitsuru IKEDA, Osamu KAKUSHO, Riichiro MIZOGUCHI
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
597-608
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
Recently much attention has been paid to the notion of "ontology" in the expectation that it can serve as the new, strong foundation of knowledge engineering. In the conventional approach to theory of knowledge, to give the operational semantics of knowledge representation has been regarded as of major importance and the analysis of contents of knowledge has been considered to be subordinate to it. To solidify the foundation of knowledge engineering, however, the many researchers, espcially in the field of knowledge sharing and reuse, has strongly felt necessity of the change of such a way of thinking. The key to the problem is to understand the essential interaction between "form" and "contents" on equal importance. This implies that deep understanding of "content" will give us new insight into design of knowledge representation. The notion of "ontology" can be key to this issue. We have investigated the property of problem solving knowledge and tried to design its ontology, that is, task ontology. The main purpose of this paper is to discuss the basic issues concerning the task ontology. It provides three major advantages as follows. (A) It provides human-friendly primitives in terms of which users can easily describe their own problem solving process. (B) The systems with task ontology can simulate the problem solving process at an abstract level in terms of conceptual level primitives. (C) It provides ontology author with an environment for building task ontology so that he/she can build a consistent and useful ontology.
View full abstract
-
Sachiyo ARAI, Kazuteru MIYAZAKI, Shigenobu KOBAYASHI
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
609-618
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
Most of multi-agent systems have been developed in the field of Distributed Artificial Intelligence (DAI) whose schemes are based on plenty of pre-knowledge of the agents' world or organized relationships among the agents. However, these kind of knowledge would not be always available. On the other hand, multi-agent reinforcement learning is worth considering to realize the cooperative behavior among the agents with littls pre-knowledge. There are two main problems to be considered in multi-agent reinforcement learning. One is the uncertainty of state transition problem which is owing to concurrent learning of the agents. And the other is the perceptual aliasing problem which is generally held in such a world. Therefore, the robustness and flexibility are essential for the multi-agent reinforcement learning toward these two problems. In this paper, we evaluate Q-learning and Profit Sharing as the method for multi-agent reinforcement learning through some experiments. We take up the Pursuit Problem as one of the multi-agent world. In the experiments, we do not assume the existence of any pre-defined relationship among agents or any control knowledge for cooperation. Learning agents do not share sensation, episodes and policies. Each agent learns through its own episodes independent of the others. The result of experiments shows that cooperative behaviors emerge clearly among Profit Sharing hunters who are not influenced by concurrent learning even when the prey has the certain escaping way against the hunters. Moreover, they behave rational under the perceptual aliasing areas. On the other hand, Q-learning hunters can not make any policy in such a world. Through these experiments, we conclude that Profit Sharing has the good properties for multi-agent reinforcement learning because or its rubustness for the change of other agents' policies and the limitation of agent's sensing abilities.
View full abstract
-
Seiichi NAKAGAWA, Minoru SAITO, Mikio MASUKATA
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
619-630
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
In this paper, we propose a method to acquire speech procesing units automatically that are suitable for automatic processing, not based on prior units like syllables or phonemes. In addition, while the units are acquired automatically, the acquisition process of concept of speech unit on human is taken into consideration, so the restriction such that "speech utterances to the same word are represented by the same unit sequence" is added. We acquired speech units using procudures of template matching, applied these units to speech recognition and found that it needs to take into consideration the variation of speech patterns. Therefore, we modelled the speech patterns corresponding to acquired units by using HMMs and higher recognition rates were obtained by training of HMMs with these units than those by the original template units. Therefore, a method to acquire units based on ergodic HMMs from an initial step was investigated in this paper. When we evaluated these units by word recognition experiments, a high recognition rate 99.5% at 216 words was obtained, with consideration of the above restriction at the both steps of acquisition process of units and registration of word dictionary (on the conditions of phoneme-like duration and 64 numbers of units). Finally, we compared the acquired units with phonetic units and found that the former is better than the latter on spoken word recognition.
View full abstract
-
Masashi SHIMBO, Toru ISHIDA
Article type: Technical paper
1998 Volume 13 Issue 4 Pages
631-638
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
The real-time algorithm LRTA enjoys an attractive property called convergence ; through the repeated problem solving trials, the problem solver will eventually identify ( or learn) an optimal path to the nearest goal. In his original LRTA paper, Korf presented a proof of convergence, but only on the assumption that the initial heuristic estimates satisfy consistency ; it was not made clear whether the convergence is retained under inconsistent heuristics, though the extension of his proof to this case is nontrivial. In this article, we establish the convergence of LRTA by a novel technique that does not rest on the consistency assumption at all. Since it is a natural extension of the proof of the completeness, it constitutes the connection between these fundamental properties of LRTA that have been discussed somewhat independently. We also compare our technique with that of Barto et al. who proved general convergence by restating LRTA as an instance of asynchronous dynamic programming.
View full abstract
-
Kenki MATSUDA, Yukio OHSAWA, Masahiko YACHIDA
Article type: Research note
1998 Volume 13 Issue 4 Pages
639-643
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
In previous travel navigation systems, the user had to input the goal and the order of places to visit. Such systems don't work if the user has nebulous demands for his/her travel. We propose a new travel navigator with a graphical dialogue interface, i. e. a two-dimensional interface. The user composes a question on that interface, then the system returns a travel plan. When the user is not satisfied with the returned plan, he/she inputs one's demands again. In this manner, the system puts the user's nebulous demands into concrete ones.
View full abstract
-
Hidetaka NAMBO, Haruhiko KIMURA, Sadaki HIROSE
Article type: Corner article
1998 Volume 13 Issue 4 Pages
644-651
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
One of the important topics in production systems is how to deal with expensive productions. Expensive productions are rules which would be required the extraordinary time and space to match (compare the left-hand side of rule against the current working memory). In this paper, we propose a match algorithm which improved the direct match algorithm for managing the working memory elements (WMEs) compressed by their attribute values. By compression of WMEs, the number of WMEs which must be examined whether the condition of rule matches or not decreases. Therefore, the frequency of match decreases, and the time for matching will be improved. We also compare the proposed algorithm to Rete algorithm and direct match algorithm, and show that the proposed algorithm is effective to deal with some class of expensive productions.
View full abstract
-
[in Japanese]
Article type: Other
1998 Volume 13 Issue 4 Pages
652-653
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
1998 Volume 13 Issue 4 Pages
654-655
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
1998 Volume 13 Issue 4 Pages
656
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
1998 Volume 13 Issue 4 Pages
657
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
1998 Volume 13 Issue 4 Pages
658
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Article type: Activity report
1998 Volume 13 Issue 4 Pages
659-665
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Article type: Activity report
1998 Volume 13 Issue 4 Pages
666-667
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Article type: Activity report
1998 Volume 13 Issue 4 Pages
668
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Article type: Activity report
1998 Volume 13 Issue 4 Pages
b001-b012
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Article type: Cover page
1998 Volume 13 Issue 4 Pages
c004
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Article type: Cover page
1998 Volume 13 Issue 4 Pages
c004_2
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Article type: Table of contents
1998 Volume 13 Issue 4 Pages
i004
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Article type: Table of contents
1998 Volume 13 Issue 4 Pages
i004_2
Published: July 01, 1998
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS