Self-organizing Rules Research Articles

Colony-specific architecture of shelter tubes by termites

Social insects build sophisticated and complex architectures such as huge nests and underground galleries based on self-organizing rules. The structures of these architectures vary widely in size and shape within a species. Some studies have revealed that the current environmental and/or social factors can cause differences in the architectures that emerge from collective building. However, little is known about the effect of colony-level variations on the architecture. Here, we demonstrate that termite colonies build colony-specific architecture using shelter-tube construction as a model system. When we divided a colony into multiple groups of individuals, groups drawn from the same colony performed similar patterns of construction, whereas groups from different colonies exhibited different patterns. The colony variations in shelter-tube construction are generally thought to reflect differences in foraging strategy, and this difference can have important fitness consequences depending on the distribution of wood resources in the environment. This is the first demonstration of colony variation in the architecture that emerges from collective behavior. Colony-specific architectural variations provide new insights into our understanding of the self-organization systems, which were previously assumed to provide each species with a species-specific construction mechanism.

Institutional diversity and local forest governance

• We investigate the effects of distinct institutional arrangements on forest conditions. • We analyze observations from 200 rural communities in Bolivia. • Institutions for rule-making, monitoring, and sanctioning affect forest conditions. • The more governance activities local users engage in, the better it is for forests. Scholarship on common-pool resource governance suggests that collective outcomes vary with the strength of the local arrangements for compliance monitoring. Following Elinor Ostrom's approach to question panaceas, we explore the possibility that there are multiple institutional designs can help sustain forests. We test this argument with data from a sample of 200 forest user groups in Bolivia and find broad empirical support for our propositions: Local monitoring can be an important predictor of forest governance performance, but focusing on monitoring alone can be misleading. Sometimes other aspects of the local governance system, such as self-organized rule making and sanctioning, are more important in explaining why some groups govern their forests more effectively than others. We also find that the more governance functions that communities decide to organize themselves, the more likely it is that local forests are sustained.

Novel self-organizing rules for retinotopic remapping

Event Abstract Back to Event Novel self-organizing rules for retinotopic remapping Markus Butz1* and Arjen Van Ooyen2 1 Ruhr-University Bochum, Institute for Neural Computation, Germany 2 VU Amsterdam, CNCR, Netherlands Retinotopic maps in the primary visual cortex are plastic even in the mature brain. Particularly after permanent changes in external input, i.e. after focal retinal lesions, maps in the cortex adapt so that neurons deprived of input (lesion projection zone, LPZ) become responsive to adjacent input representations. This 'filling-in' process is currently explained by STDP. However, STDP is a fast process while the time course of reorganization continues over weeks until one year [1]. Recent data indicates that structural plasticity (forming new synapses, breaking old ones) is involved in this reorganization [2,3] and acts very much on the same time scale. Therefore, we propose the first model investigating structural plasticity in application to cortical remapping. The model implements local activity-dependent rules for changes in the morphology of the neuron. In accordance with experimental findings, model neurons aim to maintain their electrical activity on average at a certain pre-defined set-point [4] by adapting the number of contact sites (axonal and dendritic elements). New (vacant) synaptic elements are offered to the network and connect to form synapses. The probability for synapse formation between two neurons depends on the amounts of vacant synaptic elements offered and the Euclidean distance between the two neurons. Network rewiring is therefore a reciprocal process between activity and network structure: Activity levels inside the LPZ (low) and outside the LPZ (high) locally induce the formation of axonal and dendritic elements, respectively, that in turn form synapses in a cooperative and compensatory manner leading to increasing activities in the LPZ again. The consequence of transporting activities via new synapses from the outside of the LPZ into the LPZ is an enlargement of input representation from intact areas and a sequential filling-in of the LPZ (Fig. 1)—by contrast not obtained in self-organizing maps. The novelty of the model is to generate predictions how local cellular responses lead to rewiring and remapping on an anatomical network level. Caption Figure 1: Cortical remapping emerges from new horizontal connections formed from intact areas into the LPZ. A) Colors indicate spatial input representations. White dashed circle indicates LPZ. B) New connections impinging on cells in the LPZ (white dots). Connections originating in the peri-LPZ (green), LPZ border (orange), LPZ center (blue). Horizontal and vertical bars indicate the relative position of the areas to the entire network. Figure 1 Acknowledgements MB was supported by a grant (635.100.017), awarded to AvO, of the Computational Life Sciences program of the Netherlands Organization for Scientific Research.

Self-organizing ARTMAP rule discovery

The self-organizing ARTMAP rule discovery (SOARD) system derives relationships among recognition classes during online learning. SOARD training on input/output pairs produces the basic competence of direct recognition of individual class labels for new test inputs. As a typical supervised system, it learns many-to-one maps, which recognize different inputs (Spot, Rex) as belonging to one class (dog). As an ARTMAP system, it also learns one-to-many maps, allowing a given input (Spot) to learn a new class (animal) without forgetting its previously learned output (dog), even as it corrects erroneous predictions (cat). As it learns individual input/output class predictions, SOARD employs distributed code representations that support online rule discovery. When the input Spot activates the classes dogand animal, confidence in the rule dog→animal begins to grow. When other inputs simultaneously activate classes cat and animal, confidence in the converse rule, animal→dog, decreases. Confidence in a self-organized rule is encoded as the weight in a path from one class node to the other. An experience-based mechanism modulates the rate of rule learning, to keep inaccurate predictions from creating false rules during early learning. Rules may be excitatory or inhibitory so that rule-based activation can add missing classes and remove incorrect ones. SOARD rule activation also enables inputs to learn to make direct predictions of output classes that they have never experienced during supervised training. When input Rex activates its learned class dog, the rule dog→animal indirectly activates the output class animal. The newly activated class serves as a teaching signal which allows input Rex to learn direct activation of the output class animal. Simulations using small-scale and large-scale datasets demonstrate functional properties of the SOARD system in both spatial and time-series domains.

Brain science of the mind

The human cerebral cortex contains more than one hundred billion neurons and 1014 synapses. Even without regard to the size of the genome, it can be easily deduced that a deterministic blueprint for connectivity of such an enormous number of networks is unrealistic. Existing scientific knowledge indicates that nature utilizes principal rules instead of complete deterministic descriptions to fashion a desired structure, namely, the rules of self-organization. The brain is a complex system and self-organizes based on the Markovian process. Accordingly, brain functionality can be seen as specific patterns created by self-organizing processes based on conditions defined by genes and the environment. Three categorically different systems are now recognized based on their physiological functional unit configuration. While the oldest system is made up of deterministic connectivity, the remaining two systems, namely, cerebellum and cerebrum, utilize modifiable units, often referred to as cerebellar chip and brain chip, respectively. Classical conditioning, as in the case of Pavlov’s dog, is now recognized to be based on functionality of the cerebellum and its learning unit, the cerebellar chip, which in principle works as McCulloch-Pitts neurons. The cerebrum utilizes the concept of Kohonen’s non-linear self-organizing map in the organization of the brain chip, a system that effectively creates entropy fields. In contrast to cerebellar learning which is adaptive, cerebral learning is stochastic in nature, and follows the rule known as Pólya’s Urn.

Experimental analysis of control loops with different delay times in the supply air system of a radiator test rig

The test rig of low-temperature hot-water radiator is a typical multi-input multi-output (MIMO) thermal system with large time delay. Aiming at the control of the supply air temperature and the temperature of the environmental chamber in the test rig, the paper carries out the fuzzy control experiments and PI control experiments based on adjusting the electric heater in the supply air system, in which there are two control loops with various delay time. The experiments researched two types of fuzzy control methods which are the basic fuzzy control (BFC), and the self-organizing rules fuzzy control (SORFC) with zero initial control rules. For the control loop of supply air temperature with short delay time, PI control performs better with fast convergence speed and small overshoot than the BFC and the SORFC. The convergence speed of the BFC method is lower than other two control methods. For the control loop of chamber temperature with long delay time, the SORFC has a satisfied control performance with less overshoot and stable error than the BFC and PI control.

Optimization of Self-organization Model of Electronic Connector under Fretting Wear Based on Taguchi Method

The self-organization model of an electronic connector under fretting wear is studied to explain the invalidation mechanism in fretting wear driven by integrating local rules. The electronic connector under fretting wear is the object to be studied. In order to confirm the local rules of self-organization model of fretting wear and their parameters,Taguchi method is proposed to optimize the self-organization model of an electronic connector under fretting wear. Taguchi method is applied to set up a dominant relationship between the self-organization rules and evaluation function quantitatively. That is to say,the experimental data of an electronic connector under fretting wear is regarded as the evaluation function. And Taguchi method is used to design and solve the simulation of unknown parameters in an integrated rule of the cellular automata model. Thus,from a high quality self-organization model,the invalidation mechanism of an electric connector can be expressed in high precision. Meanwhile,the study and practice on the subject shows that Taguchi method is an effective and simple tool to solve the complicated system based on the integration of a bottom-up self-organization theory and a top-down planning theory.

Iterated function systems in the hippocampal CA1

How does the information of spatiotemporal sequence stemming from the hippocampal CA3 area affect the postsynaptic membrane potentials of the hippocampal CA1 neurons? In a recent study, we observed hierarchical clusters of the distribution of membrane potentials of CA1 neurons, arranged according to the history of input sequences (Fukushima et al Cogn Neurodyn 1(4):305-316, 2007). In the present paper, we deal with the dynamical mechanism generating such a hierarchical distribution. The recording data were investigated using return map analysis. We also deal with a collective behavior at population level, using a reconstructed multi-cell recording data set. At both individual cell and population levels, a return map of the response sequence of CA1 pyramidal cells was well approximated by a set of contractive affine transformations, where the transformations represent self-organized rules by which the input pattern sequences are encoded. These findings provide direct evidence that the information of temporal sequences generated in CA3 can be self-similarly represented in the membrane potentials of CA1 pyramidal cells.

Iterated function systems in the hippocampal CA1

How does the information of spatiotemporal sequence stemming from the hippocampal CA3 area affect the postsynaptic membrane potentials of the hippocampal CA1 neurons? In a recent study, we observed hierarchical clusters of the distribution of membrane potentials of CA1 neurons, arranged according to the history of input sequences (Fukushima et al Cogn Neurodyn 1(4):305–316, 2007). In the present paper, we deal with the dynamical mechanism generating such a hierarchical distribution. The recording data were investigated using return map analysis. We also deal with a collective behavior at population level, using a reconstructed multi-cell recording data set. At both individual cell and population levels, a return map of the response sequence of CA1 pyramidal cells was well approximated by a set of contractive affine transformations, where the transformations represent self-organized rules by which the input pattern sequences are encoded. These findings provide direct evidence that the information of temporal sequences generated in CA3 can be self-similarly represented in the membrane potentials of CA1 pyramidal cells.

From Compromise to Leadership in Pigeon Homing

A central problem faced by animals traveling in groups is how navigational decisions by group members are integrated, especially when members cannot assess which individuals are best informed or have conflicting information or interests . Pigeons are now known to recapitulate faithfully their individually distinct habitual routes home , and this provides a novel paradigm for investigating collective decisions during flight under varying levels of interindividual conflict. Using high-precision GPS tracking of pairs of pigeons, we found that if conflict between two birds' directional preferences was small, individuals averaged their routes, whereas if conflict rose over a critical threshold, either the pair split or one of the birds became the leader. Modeling such paired decision-making showed that both outcomes-compromise and leadership-could emerge from the same set of simple behavioral rules. Pairs also navigated more efficiently than did the individuals of which they were composed, even though leadership was not necessarily assumed by the more efficient bird. In the context of mass migration of birds and other animals, our results imply that simple self-organizing rules can produce behaviors that improve accuracy in decision-making and thus benefit individuals traveling in groups .

MOVE-FORWARD RULES AND f-SWAP RULES APPLIED TO A COMMUNICATION PROBLEM

In a communication network, one might attempt to route calls from an origin to a destination through n paths that will be tried one by one, each having a success probability pi ∈ (0,1), i = 1,2,…,n. The order of trying is controlled by a routing table. The number of attempts made is defined as the cost of the routing table. Move-forward self-organizing rules are applied to the routing table and comparisons of expected equilibrium costs are performed when p2 = p3 = … = pn. Stationary distributions for a subset of f-swap rules are obtained for general pi's.

Sogarg: a self-organized genetic algorithm-based rule generation scheme for fuzzy controllers

This paper presents a self-organized genetic algorithm-based rule generation (SOGARG) method for fuzzy logic controllers. It is a three-stage hierarchical scheme that does not require any expert knowledge and input-output data. The first stage selects rules required to control the system in the vicinity of the set point. The second stage extends this to the entire input space, giving a rulebase that can bring the system to its set point from almost all initial states. The third stage refines the rulebase and reduces the number of rules. The first two stages use the same fitness function whose aim is only to acquire the controllability, but the last stage uses a different one, which attempts to optimize both the settling time and number of rules. The effectiveness of SOGARG is demonstrated using an inverted pendulum and the truck reversing.

Free convection schema defines the shape of the human brain: brain and self-organization

As a fundamental rule of nature, the shape of a functioning organ should never be considered accidental. Since the principal rules for a self-organizing neural network are not sufficient for determining the shape and anatomic features of the brain, these must be governed by other rules. Ontogeny of the global shape of the brain is guided by radial glial fibers. The rules defining the growth pattern of radial glial fibers, therefore, should be the rules for a self-organization for the shape of the brain. In this work, the overall shape and structural anatomic detail of the human brain was successfully simulated utilizing self-organization rules of heat convection represented by Boussinesq partial differential equations. The result suggests that radial glial fibers may grow in Markovian fashion, guided by gaseous substrates moving miniature distances according to a convection schema. The result also holds implication that the shape of the brain has specific physical meaning in the creation of a virtual sphere, which in turn ensures equivalence of columns within the entire cortex, the well-known vital fundamental of brain organization.

DISTRIBUTED PROBLEM SOLVING WITHOUT COMMUNICATION — AN EXAMINATION OF COMPUTATIONALLY HARD SATISFIABILITY PROBLEMS

In this paper, we extend and modify the ERA approach proposed in Ref. 13 to solve Propositional Satisfiability Problems (SATs). The new ERA approach involves a multiagent system where each agent only senses its local environment and applies some self-organizing rules for governing its movements. The environment, which is a two-dimensional cellular environment, records and updates the local values that are computed and affected according to the movements of individual agents. In solving a SAT with the ERA approach, we first divide variables into several groups, and represent each variable group with an agent whose possible positions correspond to the elements in a Cartesian product of variable domains, and then randomly place each agent onto one of its possible positions. Thereafter, the ERA system will keep on dispatching agents to choose their movements until an exact or approximate solution emerges. The experimental results on some benchmark SAT test-sets have shown that the ERA approach can obtain comparable results as well as stable performances for SAT problems. In particular, it can find approximate solutions for SAT problems in only a few steps. The real value of this approach is that it is a distributed asynchronous approach without any centralized control or evaluation, where the agents can cooperate to solve problems without explicit communication.

Designing Cellular Self-organization Algorithms for Multimicrocontrollers with Programmable Backup

Design of a cellular self-organization algorithm for fault-tolerant multimicrocontrollers in the form of a lattice of working and backup multiprogram microcontrollers was discussed. To program arbitrary lattice elements as backups, a model and rules of self-organization of multimicrocontrollers were determined. Structural and functional organization of the tools for retargeting the multimicrocontroller were described. Examples of executing the cellular self-organization algorithm were presented.

A Class of Robust Principal Component Vectors

This paper is concerned with a study of robust estimation in principal component analysis. A class of robust estimators which are characterized as eigenvectors of weighted sample covariance matrices is proposed, where the weight functions recursively depend on the eigenvectors themselves. Also, a feasible algorithm based on iterative reweighting of the covariance matrices is suggested for obtaining these estimators in practice. Statistical properties of the proposed estimators are investigated in terms of sensitivity to outliers and relative efficiency via their influence functions, which are derived with the help of Stein's lemma. We give a simple condition on the weight functions which ensures robustness of the estimators. The class includes, as a typical example, a method by the self-organizing rule in the neural computation. A numerical experiment is conducted to confirm a rapid convergence of the suggested algorithm.

A generalized abel's partial summation formula and its application in self-organizing systems

In this paper, Abel's partial summation formula is generalized to a two-dimensional case. The result finds an application in comparing self-organizing algorithms. It is shown that, under all request probabilities, the expected equilibrium search cost for the POS(2) rule is not less than that for the POS(3) rule, both rules being self-organizing rules for arranging records in a linear list

Applying POS(i) rules to communication problems

A spectrum of self-organizing rules including the move-to-front rule and the transposition rule are applied to the communication problem. The stationary distributions under these rules are obtained. Cost comparison between them is considered. In the special case of three paths, it is shown that the transposition rule always outperforms the move-to-front rule.

Applying POS(i) rules to communication problems

A spectrum of self-organizing rules including the move-to-front rule and the transposition rule are applied to the communication problem. The stationary distributions under these rules are obtained. Cost comparison between them is considered. In the special case of three paths, it is shown that the transposition rule always outperforms the move-to-front rule.

The Influence Function of Principal Component Analysis by Self-Organizing Rule.

This article is concerned with a neural network approach to principal component analysis (PCA). An algorithm for PCA by the self-organizing rule has been proposed and its robustness observed through the simulation study by Xu and Yuille (1995). In this article, the robustness of the algorithm against outliers is investigated by using the theory of influence function. The influence function of the principal component vector is given in an explicit form. Through this expression, the method is shown to be robust against any directions orthogonal to the principal component vector. In addition, a statistic generated by the self-organizing rule is proposed to assess the influence of data in PCA.