* 1979.3 BA Eng. in Applied Mathematics and Physics, Kyoto University * 1981.3 MA Eng. in Applied Mathematics and Physics, Kyoto University * 1986.3 Dr. Eng. in Applied Mathematics and Physics, Kyoto University
Career
* 1983.4-1986.3 Assistant Professor at Department of Applied Mathematics and Physics, Kyoto University * 1986.4-1994.3 Assistant Professor at Division of Applied Systems Science, Kyoto University * 1986.4-1987.3 Visiting Researcher at School of Computer Science, Carnegie-Mellon University * 1993.4-1994.3 Visiting Researcher at Department of Psychology, Carnegie-Mellon University * 1994.4-1998.5 Associate professor at Graduate School of Information Science, Nara Institute of Science and Technology * 1997.9-1998.2 Visiting Researcher at Santa Fe Institute * 1998.6 Professor at Department of Knowledge-based Information Engineering, Toyohashi University of Technology * 2010.4 Professor at Department of Computer Science and Engineering, Toyohashi University of Technology
Abstract: Abstract Cellular automata produce spatial patterns when specific rules for time development are given. This paper deals with an inverse problem of identifying the rules for spatial patterns given. Although only rules of one-dimensional elementary cellular automata and one-dimensional probabilistic cellular automata have shown here, the system can deal with two-dimensional one. When the rule identification has not been fully successful due to the lack of information in the spatial pattern, the system is able to give an identifiable part of the rules with a format of Wolfram’s rule number.
Abstract: Abstract We constructed a probabilistic simulator that allows all the events in population dynamics such as death, birth, mutation, and suppression/stimulation to be described by probabilistic rules. The simulator also facilitates a lattice used for expressing distribution and diversity (number of distinct strains) of quasispecies. The simulator is used to investigate the diversity threshold in HIV and T-cell interaction.
Abstract: Abstract We have proposed a self-repairing network where nodes are capable of repairing neighboring nodes by mutually copying. A critical point where faulty nodes can be eliminated has been investigated. This paper further studies the dynamics of eradicating faulty nodes by comparing the self-repairing network with mathematical epidemic models such as SIS models. It is shown that the self-repairing network, which is a probabilistic cellular automaton, can be regarded as an epidemic model in some restricted situations.
Abstract: Abstract We investigate a synchrony in a self-repairing network of autonomous agents capable of repairing mutually. In this paper, we define two models: a synchronous model and asynchronous one. They differ in the timing when the agents change their state. Computer simulations revealed that the synchronous model has a critical point, while the asynchronous one does not. We also studied a repair scheme in asynchronous model where the repaired agents in turn repair neighbor agents successively in a chain-reactive fashion. Performance of the scheme has been examined by computer simulations.
Abstract: Abstract Antibodies, among other things, are important components of the immune system. This paper proposes using the specific recognition capability exhibited by antibodies for computation, in particular, for solving the stable marriage problem, which has been studied as a combinatorial computational problem. Antibody-based computation is proposed by integrating the recognition capabilities of antibodies. The computation is carried out on an array form that is suitable not only for expressing stable marriage problems, but also for further integration to antibody microarrays.
Abstract: Abstract An application of the antibody’s flexible recognition (i.e. multi-reactivity) to antigenic epitopes to a combinatorial computing is just getting started. The present study discusses an antibody-based computation algorithm to solve a combinatorial problem: the stable marriage problem. The stable marriage problem supposes n men and n women, and each person ranks all members of the opposite sex in a strict order of preference. Under given preference lists, to detect all of “stable†n couples including no affair pairs means to solve this problem. Our algorithm replaces a man and a woman with an antigenic epitope and an antibody respectively, and re-scales a man (woman)’s preference to a woman (man) as strength of a binding affinity between an epitope to the man and an antibody to the woman. Under these settings, we demonstrate a parallel progression of immune reactions can solve the stable marriage problem.
Abstract: Many strategies, such as tit-for-tat, have been proposed in the iterated prisoner's dilemma (IPD) in which the prisoner's dilemma (PD) is carried out repeatedly with two players. A spatial version of the iterated prisoner's dilemma (SPD) has been studied, where a player at each site plays the IPD game with all the players in the neighborhood. However, the strategies studied in the SPD consider the past actions of a single opponent only. We studied spatial strategies that depend on the configuration of actions taken by all neighbors (as opposed to conventional temporal strategies). Since generosity can be considered as a spatial strategy, we first investigate the generosity required when an action error is involved. We also propose several spatial strategies that outperform many others.
Abstract: Distributed intrusion detection systems have several advantages over centralized systems, such as scalability, adaptability, and fault tolerance. A current research topic in distributed systems is self-monitoring to identify corrupted intrusion detection systems. One way of self-monitoring is for intrusion detection systems to check each other. As we describe herein, this can be done by mobile agents using an immunity-based diagnostic method modeled on idiotypic network theory. In simulations, the credibility of normal intrusion detection systems remained near 1, while it fell to about 0 for corrupted intrusion detection systems, thus enabling identification of the latter. We also confirmed what effects some parameters have on the diagnostic capability.
Abstract: Abstract A stochastic model based on the Nowak-May model for interaction between the human immunodeficiency virus (HIV) and the immune system is presented. From simulations with the stochastic model expressed on lattice spaces, an antigenic diversity threshold for HIV is shown. Mutation of HIV causes the antigenic diversity threshold to decline. However, most of the behavior in this stochastic model strongly depends on intial conditions.
Abstract: ;We propose an adaptive algorithm based on some features of the immune system (a selection-based mechanism compatible with Edelman's selectionist principle, self/nonself reference, and negative/positive selection). The algorithm proceeds in three steps: diversity generation, establishment of self-tolerance, and memorizing nonself. This algorithm may typically be used to model the system of distributed agents where the system (the self) as well as the environment (the nonself) are unknown or cannot be modeled. An agent-based architecture based on the local memory hypothesis and a network-based architecture based on the network hypothesis are discussed. The agent-based architecture is elaborated with applications to an adaptive system where knowledge about the environment is not available. An adaptive noise neutralizer is formalized and simulated for a simple plant.
Abstract: Abstract A distributed and autonomous sensor network is proposed based on the informational features of the immune network: recognition of nonself by distributed and dynamically interacting units, recognition by a simple comparison with the units themselves, dynamic propagation of activation that would lead to system-level recognition, and memory embedded as stable equilibrium states in the dynamic network. The network is explained by an illustrative example of an eight-coin puzzle: a balance must be used only three times to identify one coin with a different weight from the other seven coins. Our network also uses a dynamic structure network rather than the fixed structures used in neural networks. Simulations show that nonself (the different coin in the eight-coin puzzle, the sensor/process fault in the monitoring example) will be identified by dynamically propagating activation through the network.
Abstract: We considered a self-repair network by an autonomous and strategic repair. The network is assumed to be composed of agents with a failure rate. This paper further assumes that the failure rate changes dynamically, hence modelling a dynamic environment. When the failure rate oscillates with a fixed amplitude and cycle, computer simulations indicated that there is a threshold of not only the amplitude but the cycle. When the failure rate changes with the cycle not exceeding the threshold, the strategic repair adapts to the environment and exhibits a reasonable performance.
Abstract: A spatio-temporal generalization involves not only conventional temporal strategies that determine the player's action based on the opponent past actions but also spatial strategies based on the neighbor players' current actions and configurations. This framework allows the model to be dealt as a second order cellular automaton. With this involvement of the spatial strategies, we have observed a membrane formation which protects the cooperating clusters from being corroded by defecting intruders.
Abstract: We have developed a web-based system of presenting Sangaku. Sangaku is a way of presenting theorems and conjectures in geometry in one sheet of votive picture. Sangaku was used for temples in Japan to exchange geometrical findings, and has been studied as a traditional form of Japanese mathematics called Wasan. We developed a diagrammatic reasoning that directly uses geometric objects as primitives for reasoning, while tools called dynamic geometry software have also been developed. In developing the web-based Sangaku presentation system, we used diagrammatic reasoning which allows users to directly manipulate geometric objects. As a result, geometric findings can be presented not in a static manner of a single picture, but in a dynamic manner of operational geometric objects. Some examples of Sangaku can be seen at: http://www.sys.tutkie.tut.ac.jp/MuJapan.html.
Abstract: The post-genome era proved that DNA sequence data [11, 26] with structural and functional analysis on genes archived in many data bases can support in developing new bio-engineering technologies and can drive systemic views for biological systems. However, the post-genome era also proved that sequence data alone is not sufficient, but revealed that higher knowledge of the function of proteins is indispensable. Personalized medicine required not only sequence data, but further knowledge such as SNP (single nucleotide polymorphism) and of functioning of proteins and its deployment to interacting systems such as gene networks, giving birth of a new territory called proteome. This Chapter is organized as follows: Sect. 2 focuses on the preliminary problem of whether recognition is indeed needed, focusing on the specific task of abnormal state eradication on a simple network. Section 3 addresses the problem of networked recognition that involve action counterpart, hence agents can not only recognize but also be recognized. Section 4 further introduces adaptation by assuming agents can not only reproduce but also mutate in the receptor counterpart. Section 5 considers arrayed recognition, which is the very first step, even before networked recognition; however, it assumes specific recognition capability of antibody-antigen recognition.
Abstract: It is well known that mutations occur at high rate at the replication of HIV genome. Mutations on the gene encoding requisite proteins for infection to host cells deprive ability to infect to them, thus such mutations are fatal for HIV population. We have extended Nowak and May’s HIV dynamical model into more realistic one by considering the fatal mutations of HIV and an evolution of a mutation rate. Through analyses of the model, we have derived a critical condition controlling an onset of AIDS. The most significant consequence from the derived condition is to have clarified the antigenic diversity threshold theory proposed by Nowak and May succeeds in a characterless and homogeneous HIV population; the theory does not in an inhomogeneous HIV population.
Abstract: Complex systems have provided not only an analytic view that computational intelligence could be attained at a critical point (edge of chaos) where a phase transition takes place, but also a synthetic view that computational intelligence could be embedded in the field where an open and evolutionary environment for sel.sh agents will lead to collective phenomena. In the synthetic view, using complex systems themselves for intelligent systems, such as DNA computing (we focus on immunity-based computing in another Chapter of this volume), grid computing, and parasitic computing, is another important paradigm. This Chapter investigates the first step towards embedding computational intelligence in the Internet field by selfish agents, namely, whether selfish agents can ever cooperate and converge on some tasks. Selfish routing and task allocation have been studied extensively in the computational game community, but can intelligent tasks be done or can agents ever take care of themselves in the first place? We first pose the problem of self maintenance in an agent population, and then use a game theoretic approach to test whether cooperation would occur or under what conditions cooperation will occur.
Abstract: Our immunity-based anomaly detection system aims to detect anomalous behavior of users on a computer. To improve the detection accuracy, we introduced the framework of dynamically updating profiles into our system. Our system enables agents to update not only self profiles, but also nonself profiles. Briefly, our system enables agents to adapt to new behavior of the original users and of others. The receiver operating characteristic (ROC) analysis of our system indicated that the updating of both profiles markedly decreased both the false alarm rate and the missed alarm rate.
Abstract: In an information network composed of selfish agents pursuing their own profits, undesirable phenomena such as spam mail occur if the profit sharing and other game structures permit such equilibriums. This note focuses on applying the spatial Prisoner's Dilemma to control a network of selfish agents by allowing each agent to cooperate or to defect. Cooperation and defection respectively correspond to repair (using the self resource) and not repair (thus saving the resource) in a self-repair network. Without modifying the payoff, the network will be absorbed into the state where all the agents become defectors and abnormal. Similarly to kin selection, agents favor survival of neighbors in organizing these two actions to prevent the network from being absorbed if payoffs are measured by summing all the neighboring agents. Even with this modification, the action organization exhibits spatial and temporal adaptability to the environment.
Abstract: The spatial Prisoner's Dilemma is divided into two stages: strategy selection and action (cooperation/defection) selection. This renewal allows a spatiotemporal strategy that determines the player's next action based not only on the adversary's history of actions (temporal strategy) but also on neighbors' configuration of actions. Several space-time parallelisms and dualisms would hold in this spatiotemporal generalization of strategy. Among them, this note focuses on the generosity (how many defections are tolerated). A temporal strategy involving temporal generosity, such as Tit for Tat (TFT), exhibited good performance such as noise tolerance. We report that a spatial strategy with spatial generosity can maintain a cluster of cooperators by forming a membrane that protects against defectors. The condition of membrane formation can be formulated with the spatial generosity exceeding a certain threshold determined by the number of neighborhoods.
Abstract: For the equations of population dynamics, this note presents three symmetries: a coordinate symmetry, an additive symmetry and an exchange symmetry. Among them, additive symmetry is a new one that should be held in equations of population dynamics particularly those for quasispecies. The distinguishability between species is also stressed to obtain the stability condition of 2-dim Lotka-Volterra model that satisfies the additive symmetry.
Abstract: When Cooperate and Defect of the Spatial Prisoner’s dilemma are corresponded to repair and not repair in a self-repair network, a major problem is that agents stuck at a Nash equilibrium of mutual defection, and the network ended in all abnormal states. To resolve the problem, we have studied game theoretic regulation schemes of selfish agents. Payoff for each agent is modified to include not only its own resources left but all the resources of the neighbor agents.
Abstract: This paper discusses when selfish agents begin to cooperate instead of defect, taking a specific task of maintenance of themselves. The network cleaning problem where a collection of agents capable of repairing other agents by overwriting its content can clean the collection will be discussed. With this problem, cooperate corresponds to repairing other agents and defect to not repair. Although both defect is a Nash equilibrium: no agent is willing to repair others when only the repair cost is involved in the payoff, agents may cooperate with each other when system reliability is also incorporated in the payoff and with certain conditions satisfied. Incentive for cooperation will be stronger when further system wide criterion such as availability is involved in the payoff.
Abstract: The complex network theory has lately drawn considerable attention, because it can successfully describe the properties of many networks found in nature. In this paper, we perform the immunity-based diagnostic model on complex networks, namely, small-world networks and scale-free networks. For the distributed diagnosis, the number of nodes on large-scale networks has little effect on the diagnosis capability. In addition, some results show the performance of the diagnosis model depends on the number of links and the average path length. The results can give us valuable knowledge to incorporate the immunity-based diagnosis into real complex networks.
Abstract: As a theory on a deciding factor leading to AIDS, â€Âantigenic diversity threshold theory†has been offered by Nowak et al. This theory mentions AIDS develops when a number of mutant strains from a single HIV swells through mutations over a fixed critical number. Mutation is mainly due to transcription errors by a reverse transcriptase. Existing HIV models assume the transcription error probability (i.e. mutation probability of HIV) is constant. However it being considered the reverse transcriptase builds its gene into HIV genes as its operand, it would be natural to guess the transcription error probability varies at each transcription. Hence, this study proposes a HIV dynamical model factoring into the mechanism of HIV mutating its mutation probability. Along with earlier studies this study discusses the subject of the antigenic diversity threshold and demonstrates the proposed model has a threshold with the minimal mutation probability influencing development of AIDS.
Abstract: Antigenic diversity threshold theory by Nowak and May proposes there will be a threshold of diversity for HIV strains in a population dynamics with asymmetric interaction between specialist (T-cells) and generalists (HIVs). This paper revisits the threshold condition from a symmetric point of view. We consider how stability condition for HIVs will be altered when the asymmetry of specialists/generalists is relaxed (made symmetrical). Diversity measure is also considered from a symmetric point of view involving a distance between parameters characterizing strains.
Abstract: This paper focuses on anomaly detection in user behavior. We present a review of our immunity-based anomaly detection system, and propose a framework of the immunity-based anomaly detection system with a new mechanism of diversity generation. In the framework, each computer on a LAN generates diverse agents, and the agents generated on each computer are shared with all other computers on the LAN. The sharing of agents contributes to their diversity. In addition, we propose an evaluation framework of immunity-based anomaly detection, which is capable of evaluating the differences in detection accuracy between internal and external malicious users.
Abstract: This paper reports on the construction and use of a dynamic relational network that has been studied as an immunity-based system based on the concept of immune networks. The network is constructed by an immunological algorithm that tunes the network not to react to normal sensor data, but to react to abnormal data thereafter. The tuning is not straightforward, since even normal sensor data involve many situations such as accelerating phase and cruise phase; on-road and off-road; and running at altitude. A case study on sensor systems for the combustion control system of an automobile engine is presented.
Abstract: This paper reports a multiagent approach to a basic model inspired by the asymmetric war between HIV and T-cells. The basic model focuses on the asymmetric interaction between two types of agents: Virus Agents (abstracted from HIV) and Immune Agents (abstracted from T-cells). Virus Agents and Immune Agents, characterized respectively as "generalists" and "specialists", may be compared with asymmetric wars between computer viruses and antivirus programs, between guerrillas and armed forces, and so on. It has been proposed that antigenic diversity determines the war between HIV and T-cells. We also formalize the diversity of "generalists" that would determine whether generalists or specialists won. The multiagent simulations also suggest that there is a diversity threshold over which the specialist cannot control the generalist. In multiagent approaches, two spaces, Agent Space and Shape Space, are used to observe not only the spatial distribution of agent populations but also the distribution of antigenic profiles expressed by a bit string.
Abstract: We have applied our previous immunity-based system to anomaly detection for network traffic, and confirmed that our system outperformed the single-profile method. For internal masquerader detection, the missed alarm rate was 11.21% with no false alarms. For worm detection, four random-scanning worms and the simulated metaserver worm were detected with no missed alarms and no false alarms, while a simulated passive worm was detected with a missed alarm rate of 80.57%.
Abstract: In our previous studies, the immunity-based diagnostic model has been used by stationary agents in linked networks or by mobile agents on wired computer networks. We have not yet analyzed the performance of the diagnosis in wireless network where agents can move freely. In this paper, the diagnosis is applied to static and mobile sensor nodes in a 2-dimensional lattice space for wireless sensor network. Some simulation results show the strategy of going straight in the different direction can have the best detection rate. In addition, when the fraction of mobile nodes is changed, the transitions of the detection rate for the migration strategies are different.
Abstract: We deal with a problem of cleaning up a contaminated network by mutual copying. This problem involves not only an aspect of "the double-edged sword" where copying could further spread contamination but an aspect of mutual cooperation where resource consuming copying could be left for others. The framework of "prisoner's dilemma" has been applied, aiming at emergence of appropriate copying strategies in an adaptive manner to the network environment.
Abstract: This paper reports a critical phenomenon in a self-repair network by mutual copying. Extensive studies have been done on critical phenomena in many fields such as in epidemic theory and in percolation theory with an effort of identification of critical points. However, from the viewpoints of cleaning up a network by mutual copying, critical phenomena have not much studied. A critical phenomenon has been observed in a self-repair network. Self-repairing by mutual copying is "the double-edged sword" that could cause outbreaks with inappropriate parameters, and careful investigations are needed.
Abstract: In our previous research, we proposed a new approach for mutual repairing system using the immunity-based diagnostic mobile agents. However, we have not yet analyzed the performance by more detailed simulations and/or mathematical models. In this paper, the approach using the immunity-based diagnostic mobile agent is compared with other approaches using majority vote model or host-to-host communication. Some results show that the immunity-based diagnostic mobile agents can repair more abnormal hosts than the other methods. We also address a mathematical model to explain a phase transition.
Abstract: In recent years, self-repair has attracted much attention again for fault tolerance in distributed computer network system. In our previous research, abnormal units (host computers and mobile agents) in distributed system can be identified by the immunity-based diagnostic model. The model calculates the credibility of each unit. In this paper, based on the credibility, some units try to self-repair, that is, replace their data with data received from other units. We propose three different repairing conditions: majority vote without credibility, credibility comparison and credibility threshold. Some simulations evaluate the effectiveness of these conditions changing the numbers of initial abnormal hosts.
Abstract: A new mechanism for generating agents, modeled on the immune system, has been incorporated into our previous immunity-based system for detecting masqueraders. Akin to T cells in the thymus, the agents capable of recognizing self from nonself, or legitimate users from masqueraders, are positively selected, while those that cannot are annihilated. In our experiments, our new system was better than our previous system at detecting external masqueraders. We also discuss the diversity and specificity of the agents.
Abstract: In distributed intrusion detection systems, self-monitoring is a difficult problem. One possibility is that each intrusion detection system is checked periodically by others. In our previous research for the self-monitoring, we proposed mutual tests between intrusion detection system and mobile agent using an immunity-based diagnosis. However, we have not clarified advantages and disadvantages of using mobile agent. In this paper, we compare the approach using mobile agent with that using host-to-host communication. Some simulation results show that in comparison with the host-to-host communication, mobile agents need twice as much time to detect corrupted intrusion detection systems, whereas a smaller number of mobile agents can identify correctly.
Abstract: An immunity-based approach that utilizes multiple profiles for detecting masqueraders in UNIX-like systems has been developed and evaluated. The approach was independent of the profile construction method. Experimental results can be summarized as follows: 1) the present approach outperformed a number of previous approaches; 2) performance was almost independent of the number of accounts when the number of accounts exceeded 10; 3) the addition of profiles enhanced performance.
Abstract: This paper explores a constructive systems approach to understand the immune system, starting from antibodies which are major units bearing specific recognition of the adaptive immune system. The exploration proceeds in stages: arrayed recognitions; networked recognition/actions; and diversified recognition/actions. System theoretic aspects of the immune system will be discussed with respect to possible application of immunity-based problem solving
