-
Toru ISHIDA
Article type: Preface
2000 Volume 15 Issue 6 Pages
939
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Hideaki TAKEDA, Kiyoshi IZUMI, Hajime KITA
Article type: Cover article
2000 Volume 15 Issue 6 Pages
940
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Kiyoshi IZUMI, Kazuhiro UEDA
Article type: Special issue
2000 Volume 15 Issue 6 Pages
941-950
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
In this paper we introduce artificial market studies, review their methods and results, and discuss their future directions. First, we give a definition of the artificial market studies and identify their features. According to these features, them, we classify some artificial market studies into three categories and review their typical studies. Next, we discuss frequent criticism about artificial market studies. Finally, we show some future directions of artificial market studies and their requirements.
View full abstract
-
Yoshinori SHIOZAWA
Article type: Special issue
2000 Volume 15 Issue 6 Pages
951-957
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Aki-hiro SATO, Hideki TAKAYASU
Article type: Special issue
2000 Volume 15 Issue 6 Pages
958-965
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Takao TERANO, Setsuya KURAHASHI
Article type: Special issue
2000 Volume 15 Issue 6 Pages
966-973
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Hiroshi SATO, Masao KUBO, Rikiya FUKUMOTO, Yasuo HIROOKA, Akira NAMATA ...
Article type: Special issue
2000 Volume 15 Issue 6 Pages
974-981
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Hiroshi DEGUCHI, Kiyoshi IZUMI, Yoshinori SHIOZAWA, Hideki TAKAYASU, T ...
Article type: Special issue
2000 Volume 15 Issue 6 Pages
982-989
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Katsumi NITTA
Article type: Cover article
2000 Volume 15 Issue 6 Pages
990
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
991
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
992
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
993
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
994
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
995
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Lobo Oscar ORTEGA
Article type: Special issue
2000 Volume 15 Issue 6 Pages
996
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
Propositional learning algorithms successfully extract concepts from data expressed thru attributes and propositions on the attribute values. However, When the concepts acquired by propositional learning algorithms are to be used at later stages, on previously unseen instances from the same designated domain, there is no absolute confidence that the concept learned will fit all the new instances. One of the reasons behind the failure of propositional learning to achieve onehundred percent accuracy on unseen objects, is that actual domains show deficiencies, like missing values, in the quality of the available data. When missing values occur in the data, the propositional learning algorithms produce concepts that classify unseen objects with less accuracy than the classification accuracy obtained by concepts learned from complete data. This dissertation formulates a new mechanism for missing value estimation which can help to reduce the error rate of the learned concepts. The contribution formulated in this dissertationis a cascaded and ordered construction of the decision trees employed to fill the missing values. This variation on the way to obtain what is hereby named Attribute Trees was empirically shown to improve the accuracy of the decision tree learning algorithm on the completed data, while holding the computational cost to an acceptable level. Though the method was successful on increasing accuracy in some domains with artificially introduced missing values, more extensive experimentation revealed that the improvement on accuracy performance is not the general case. Thus, this dissertation additionally tries to identify the reasons for the lack of general applicability of the proposed method. A new analytical tool called mutual information analysis is defined and used for identifying domains that are compatible with the attribute trees approach for filling missing values. As a result, the sort of data for which it is worth to apply the new method was determined.
View full abstract
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
997
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
998
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Christoph NEUMANN
Article type: Special issue
2000 Volume 15 Issue 6 Pages
999
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1000
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1001
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1002
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1003
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1004
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Santi SAEYOR
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1005
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
Web repositories are exposed in form of a huge amount of heterogeneous information sources. The information is overwhelming the users. Besides the target information itself, the changes upon the previously released information are significant and worth being notified to those who perceived the out of date information as soon as possible. The changes made in Web repositories occur at unpredictable rates. Unfortunately, stock type information source has no means to inform its prospective users about the changes. While the stock type information source occupies a large percentage of sources on the Web, it is necessary to have a system that monitors changes on the Web, and provides comprehensive presentation to the prospective users. In this thesis, the author proposes a mechanism that provides change monitoring and presentation service for a large group of users based on coalition among service agents. The service agents keep improving the overall utilization factor by several schemes based on the decision made by game analysis. The author apply a cost model to the service mechanism in order to study the cooperative behavior of the service agents. The reduction of cost is designed to comply with the level of cooperation among service agents. As a result, when the proposed monitoring service is used among users, the Web repositories will virtually turn into flow type information system where the information of changes automatically flows to the users who are interested. This lets the information providers update their information pages and be confident that the pu to date information will be delivered to the potential users shortly after the actions.
View full abstract
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1006
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1007
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1008
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1009
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1010-
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Special issue
2000 Volume 15 Issue 6 Pages
1011
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Yasuyuki SUMI
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1012-1026
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Seishi OKAMOTO, Minoru SEKIGUCHI, Kazuo MISUE, Fumihito NISHINO
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1027-1034
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese], [in Japanese], [in Japanese], [in Japanese]
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1035-1042
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
Yasuyuki NISHIOKA, Toshiyuki FUJITA
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1043-1051
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
We consider a system of autonomous distributed agents which make decisions based on their payoff information. In the case that there are some interference caused by task precedence, resource conflicts and so on, realization of cooperation among agents should be an important factor to increase efficiency of the whole system. In this paper we discuss such a problem in the framework of noncooperative game theory, and propose a method of realizing situations in which agents perform cooperatively without damaging their autonomy. We change payoff information indirectly by controlling evaluation parameters for each agent. Numerical results suggest that there is an efficient way of deciding evaluation parameters according to the characteristics of agents.
View full abstract
-
Akihiro INOKUCHI, Takashi WASHIO, Hiroshi MOTODA, Kohei KUMAZAWA, Naoh ...
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1052-1063
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
In the field of data mining, much attention has been paid to the Basket Analysis. Basket Analysis derives frequent itemsets from database, but its ability of mining is limited to transaction data consisting of items. Some approaches to discover characteristic graph structure from a set of given graph structured data have been researched in the field of inductive machine learning. However, most of them are not suitable for the applications which require to search all frequent graph patterns in the large amount of data. In this paper, we propose a novel principle and its algorithm that derive the characteristic patterns which frequently appear in massive graph structured data. Our algorithm can derive all frequent patterns from both the directed and undirected graph structured data having loops, multiple types of nodes and links.
View full abstract
-
Miyuki KOSHIMURA, Ryuzo HASEGAWA
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1064-1073
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
We present a method for embedding two mechanisms, which eliminate redundant inferences caused by case-splitting, into a model generation theorem prover. One is proof condensation which eliminates unnecessary subproofs by calculating their relevancy to the proof. Another is folding-up which eliminates duplicate subproofs by using lemmas extracted from the proof. These two mechanisms are achieved by extracting literals that have contributed to the proof. We have implemented the method on a UNIX workstation and evaluated its effects for some typical problems taken from the TPTP problem library. Experimental results show that the method has the effect of eliminating redundant proofs.
View full abstract
-
Fumihiko YAMAGUCHI, Masakazu NAKANISHI
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1074-1080
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
Linear logic is introduced by Girard, and is sensitive to quantity. Quantitative rules are obtained by induction in this logic. This paper proposes quantitative semantics for linear logical formulas, and proposes a method of induction. Soundness and completeness of this method with respect to the semantics are also shown. The semantics of linear logical formula is given as a set whose elements are multi-set. Multi-set is a set whose elements can repeatedly appear, i.e.the number of element is sensitive. Quantity is represented in this manner. Entailment relation between formulas is given as inclusion relation between the sets which are the semantics of these formulas. Our logic system is a sub-system of full propositional linear logic. This system is sound with respect to the proposed semantics. The method of induction from positive only example is proposed. In linear logic, normal form of a formula, such as clause in predicate logic, is hard to introduce. Thus, the inductive operations are defined by substitution of sub formulas. Soundness of the inductive method is proved straightforwardly from soundness of the logic system. Positive only example allws that the model of hypothesis has too much elements. Thus, minimal hypothesis is introduced which is defined as a hypothesis whose proper subset connot be a hypotheses. Finally completeness of the inductive method with respect to minimal hypothesis is proved.
View full abstract
-
Nobuhiro YUGAMI, Seishi OKAMOTO
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1081-1088
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
Extracting classification rules from pre-classified instances is an important field of knowledge discovery. Typically, enormous number of classification rules are embedded in the given instances and a discovery system has to extract only useful rules effectively. There are many types of useful rules depending on what the discovered rules are used for and this paper deals with discovery of an exception rule in which each condition in its bodies has no or negative correlation with its conclusion (class label) but the combination of the conditions predicts the conclusion accurately. Exception rules are important because such hidden relationships may be unknown, new knowledge for human experts. In this paper, we first propose a new criterion that judges whether a body of a rule to be evaluated is realy required to predict its conclusion by comparing its accuracy with those of more general rules. We also propose a discovery algorithm, IIS, to extract exception rules with respect to the above criterion from training instances represented by numeric attributes. IIS effectively searches exception rules by iteratively specializing possible intervals of attributes in a rule's body. The empirical results with UCI repository show that IIS is sufficiently fast to extract exception rules with three or less attributes in their bodies.
View full abstract
-
Jun MIURA, Yoshiaki SHIRAI
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1089-1096
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
This paper proposes a novel method to schedule parallel execution of planning and action of a mobile robot. The method considers the following two types of parallelism. (1)acting while planning: if a partial planning result can be used to determine feasible actions, such actions can be executed while the planning process is still going. (2)planning while acting: if the result of the current action is predictable, planning for the next action can start in advance of the completion of the current action. The proposed method uses the notion of planning-action consistency to guide the scheduling. The method has been successfully applied to a mobile robot navigation problem under sensor uncertainty.
View full abstract
-
Naruaki TOMA, Satoshi ENDO, Koji YAMADA
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1097-1106
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
The immune system is one of the adaptive biological systems whose functions are to identify and to eliminate foreign material. Applying this system to the engineering problems, these processes are regarded as an effective search and optimization. In this paper, we propose an optimization algorithm, called Adaptive Memorizing Immune Algorithm(AMIA), based on an immune system with two memory mechnisms. The one, primary memory mechnism is msed 1)to memorize antibodies that have useful characteristics as solutions and 2)to improve ability of local searching. The other, secondary memory mechanism is used to restrain re-searching of memorized solutions. Some computer simulations are designed to investigate the performance of the immune algorithm. The results of simulations represent that the immune algorithm shows good performance for multimodal functions.
View full abstract
-
Takashi WASHIO, Hiroshi MOTODA
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1107-1116
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
A novel and generic theory is formulated to characterize the structure of simultaneous equations. Based on the theory, an algorithm is developed to discover the structure of simultaneous equations governing an objective process through experiments, and the algorithm is implemented to the "Simultaneous Structure Finder(SSF)"program. SSF combined with another system SDS which discovers a quantitative formula of a complete equation derives a quantitative model of simultaneous equations reflecting scientfic laws underlying the objective process. The power of the proposed approach comes from the use of "minimal complete subsets" contained in any simultaneous equation set. They can be identified through controlled experiments on the objective process. This novel definition nicely characterizes the structure of simultaneous equations, and provides a measure to efficiently prune unnecessary search to identify simulaneous structure of an equation set. In this paper, the theoretical foundations of the structure identification and the algorithm of SSF are presented, and its efficiency and practicality are demonstrated and discussed with large scale working examples. Since the conventional equation discovery systems could not find such simultaneous equations in an efficient manner, this work is expected to provide a new research field on the scientific equation discovery.
View full abstract
-
Wataru SUNAYAMA, Yukio OHSAWA, Masahiko YACHIDA
Article type: Technical paper
2000 Volume 15 Issue 6 Pages
1117-1124
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
In recent years, Web pages have become the most available resource of information. However, it is hard to express a user interest by words, so people are hard to access those information by using search engines. In this paper, we suggest a system which supplies keywords used in real Web pages. This system aims at making up a lack of user's knowledge and filling the gap between user's knowledge and real information. Along with these, a two-dimensional search interface is supplied to a user. In this interface, keywords are classified by a user interest and are arranged with expressing the relation between a user interest and Web pages cleary. As a result, a user can make his/her own interest concrete and can get disirable pages quickly.
View full abstract
-
Lifeng HE, Yuyan CHAO, Tsuyoshi NAKAMURA, Hidenori ITOH
Article type: Research note
2000 Volume 15 Issue 6 Pages
1125-1129
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
This paper presents an improvement of theorem prover A-SATCHMORE. Considering every consequent atoms of non-Horn clauses in a given clause set are derivable, A-SATCHMORE may suffer from a potential explosion of the search space when some of such consequent atoms are precisely underivable. In this paper, we introduce how to exclude those underivable consequent atoms of non-Horn clauses from our computation.
View full abstract
-
[in Japanese]
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1130-1133
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese], [in Japanese]
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1134-1137
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1138-1140
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1141
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1142
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1143
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS
-
[in Japanese]
Article type: Corner article
2000 Volume 15 Issue 6 Pages
1144
Published: November 01, 2000
Released on J-STAGE: September 29, 2020
MAGAZINE
FREE ACCESS