Associative search network: A reinforcement learning associative memory

An associative memory system is presented which does not require a teacher to provide the desired associations. For each input key it conducts a search for the output pattern which optimizes an external payoff or reinforcement signal. The associative search network (ASN) combines pattern recognition and function optimization capabilities in a simple and effective way. We define the associative search problem, discuss conditions under which the associative search network is capable of solving it, and present results from computer simulations. The synthesis of sensory-motor control surfaces is discussed as an example of the associative search problem.     

Cluster associative memory formation in a three-layer network of phase oscillators

A three-layer network model of oscillatory associative memory is proposed. The network is capable of storing binary images, which can be retrieved upon presenting an appropriate stimulus. Binary images are encoded in the form of the spatial distribution of oscillatory phase clusters in-phase and anti-phase relative to a reference periodic signal. The information is loaded into the network using a set of interlayer connection weights. A condition for error-free pattern retrieval is formulated, delimiting the maximal number of patterns to be stored in the memory (storage capacity). It is shown that the capacity can be significantly increased by generating an optimal alphabet (basis pattern set). The number of stored patterns can reach values of the network size (the number of oscillators in each layer), which is significantly higher than the capacity of conventional oscillatory memory models. The dynamical and information characteristics of the retrieval process based on the optimal alphabet, including the size of “attraction basins“ and the input pattern distortion admissible for error-free retrieval, are investigated.     

Dynamics control of semantic processes in a hierarchical associative memory

 A neural mechanism for control of dynamics and function of associative processes in a hierarchical memory system is demonstrated. For the representation and processing of abstract knowledge, the semantic declarative memory system of the human brain is considered. The dynamics control mechanism is based on the influence of neuronal adaptation on the complexity of neural network dynamics. Different dynamical modes correspond to different levels of the ultrametric structure of the hierarchical memory being invoked during an associative process. The mechanism is deterministic but may also underlie free associative thought processes. The formulation of an abstract neural network model of hierarchical associative memory utilizes a recent approach to incorporate neuronal adaptation. It includes a generalized neuronal activation function recently derived by a Hodgkin-Huxley-type model. It is shown that the extent to which a hierarchically organized memory structure is searched is controlled by the neuronal adaptability, i.e. the strength of coupling between neuronal activity and excitability. In the brain, the concentration of various neuromodulators in turn can regulate the adaptability. An autonomously controlled sequence of bifurcations, from an initial exploratory to a final retrieval phase, of an associative process is shown to result from an activity-dependent release of neuromodulators. The dynamics control mechanism may be important in the context of various disorders of the brain and may also extend the range of applications of artificial neural networks. Received: 19 April 1995/Accepted in revised form: 8 August 1995     

Dynamics control of semantic processes in a hierarchical associative memory

A neural mechanism for control of dynamics and function of associative processes in a hierarchical memory system is demonstrated. For the representation and processing of abstract knowledge, the semantic declarative memory system of the human brain is considered. The dynamics control mechanism is based on the influence of neuronal adaptation on the complexity of neural network dynamics. Different dynamical modes correspond to different levels of the ultrametric structure of the hierarchical memory being invoked during an associative process. The mechanism is deterministic but may also underlie free associative thought processes. The formulation of an abstract neural network model of hierarchical associative memory utilizes a recent approach to incorporate neuronal adaptation. It includes a generalized neuronal activation function recently derived by a Hodgkin-Huxley-type model. It is shown that the extent to which a hierarchically organized memory structure is searched is controlled by the neuronal adaptability, i.e. the strength of coupling between neuronal activity and excitability. In the brain, the concentration of various neuromodulators in turn can regulate the adaptability. An autonomously controlled sequence of bifurcations, from an initial exploratory to a final retrieval phase, of an associative process is shown to result from an activity-dependent release of neuromodulators. The dynamics control mechanism may be important in the context of various disorders of the brain and may also extend the range of applications of artificial neural networks.     

An associative memory that can form hypotheses: a phase-coded neural network

 Nonlinear associative memories as realized, e.g., by Hopfield nets are characterized by attractor-type dynamics. When fed with a starting pattern, they converge to exactly one of the stored patterns which is supposed to be most similar. These systems cannot render hypotheses of classification, i.e., render several possible answers to a given classification problem. Inspired by von der Malsburg’s correlation theory of brain function, we extend conventional neural network architectures by introducing additional dynamical variables. Assuming an oscillatory time structure of neural firing, i.e., the existence of neural clocks, we assign a so-called phase to each formal neuron. The phases explicitly describe detailed correlations of neural activities neglected in conventional neural network architectures. Implementing this extension into a simple self-organizing network based on a feature map, we present an associative memory that actually is capable of forming hypotheses of classification. Received: 6 December 1993/Accepted in revised form: 14 July 1994     

Towards understanding of the cortical network underlying associative memory

Declarative knowledge and experiences are represented in the association cortex and are recalled by reactivation of the neural representation. Electrophysiological experiments have revealed that associations between semantically linked visual objects are formed in neural representations in the temporal and limbic cortices. Memory traces are created by the reorganization of neural circuits. These regions are reactivated during retrieval and contribute to the contents of a memory. Two different types of retrieval signals are suggested as follows: automatic and active. One flows backward from the medial temporal lobe during the automatic retrieval process, whereas the other is conveyed as a top-down signal from the prefrontal cortex to the temporal cortex during the active retrieval process. By sending the top-down signal, the prefrontal cortex manipulates and organizes to-be-remembered information, devises strategies for retrieval and monitors the outcome. To further understand the neural mechanism of memory, the following two complementary views are needed: how the multiple cortical areas in the brain-wide network interact to orchestrate cognitive functions and how the properties of single neurons and their synaptic connections with neighbouring neurons combine to form local circuits and to exhibit the function of each cortical area. We will discuss some new methodological innovations that tackle these challenges.     

Modeling the acquisition of counting with an associative network

In the acquision of counting by children, there are three interesting phenomena (Fuson et al. 1982): (1) the number word sequence produced by children can be divided into three distinct portions, called the conventional, stable nonconventional, and unstable portions; (2) irregular number words such as fifteen are omitted more often than regular ones such as fourteen, sixteen, and seventeen; and (3) initially the number word sequence is in a recitation form, rather than in the form of an associative chain of separable serial elements. Our paper at first analyzes these phenomena from the viewpoint of associative memory by assuming the number word sequences are made up of many associative relationships between the number words. This assumption is not contradictory to the third phenomenon described above, because the associative relationships are not confined only to those between the serial number words. On the basis of these anaylses, an associative network model, HAPS proposed by one of the authors (Hirai 1983), is extended so that it can mimic some aspects of the learning of sequence which involves the above three phenomena. The learning and production of sequence by the network are simulated on a digital computer, and the results show that the three phenomena can be observed in the performance of the network.     

Adaptive control for mimo uncertain nonlinear systems using recurrent wavelet neural network

Recurrent wavelet neural network (RWNN) has the advantages such as fast learning property, good generalization capability and information storing ability. With these advantages, this paper proposes an RWNN-based adaptive control (RBAC) system for multi-input multi-output (MIMO) uncertain nonlinear systems. The RBAC system is composed of a neural controller and a bounding compensator. The neural controller uses an RWNN to online mimic an ideal controller, and the bounding compensator can provide smooth and chattering-free stability compensation. From the Lyapunov stability analysis, it is shown that all signals in the closed-loop RBAC system are uniformly ultimately bounded. Finally, the proposed RBAC system is applied to the MIMO uncertain nonlinear systems such as a mass-spring-damper mechanical system and a two-link robotic manipulator system. Simulation results verify that the proposed RBAC system can achieve favorable tracking performance with desired robustness without any chattering phenomenon in the control effort.     

Distributed sensor networks: a cellular nonlinear network perspective

Large-scale networks of integrated wireless sensors become increasingly tractable. Advances in hardware technology and engineering design have led to dramatic reductions in size, power consumption, and cost for digital circuitry, and wireless communications. Networking, self-organization, and distributed operation are crucial ingredients to harness the sensing, computing, and computational capabilities of the nodes into a complete system. This article shows that those networks can be considered as cellular nonlinear networks (CNNs), and that their analysis and design may greatly benefit from the rich theoretical results available for CNNs.     

The control of re-invasion by rats of part of a sewer network

Re-infestation by rats (Rattus norvegicus Berk.) of a section of a sewer network was studied by recording bait consumption in the manhole chambers for periods up to 18 weeks after each of three attempts to rid the section of rats. The mean rate of increase of the rat population was of the order of 20 % per week both when the section was surrounded by a permanently poison-baited zone as a barrier against re-invasion from the connected, still heavily-infested sewers, and when this barrier was absent. When, however, infestation in the connected sewers was reduced by poisoning, and the poison-bait barrier was maintained, the mean weekly rate of increase in the section declined to 4·4% and could be attributed substantially to breeding. It is concluded that the type of poison-bait barrier used in the present trials is insufficient to prevent invasion of one part of a sewer network from another if the population density in the latter is high.     

Cooperative control of tumor suppressor genes by a network of oncogenic microRNAs

Individual microRNAs (miRNAs) have been implicated as oncogenes in experimental cancer models and their expression may affect clinical outcomes. To gain a more comprehensive view of miRNA action in leukemia, we analyzed miRNA expression patters in T-cell leukemia ALL (T-ALL) and cross-referenced the results with an unbiased genetic screen and computational analyses.1 We found that multiple microRNAs contribute to leukmogenesis and act as multi-targeted regulators of several tumor suppressor genes. The oncomirs form a network of overlapping and partially redundant interactions that stabilize the malignant phenotype though coordinate repression of cellular failsafe programs. The emerging network pattern of oncomir action is distinct from the notion of single oncogenic 'driver' mutation. We will discuss experimental, diagnostic and therapeutic implications of this concept of miRNA action in cancer.     

Inferring network connectivity by delayed feedback control

We suggest a control based approach to topology estimation of networks with N elements. This method first drives the network to steady states by a delayed feedback control; then performs structural perturbations for shifting the steady states M times; and finally infers the connection topology from the steady states' shifts by matrix inverse algorithm (M = N) or l(1)-norm convex optimization strategy applicable to estimate the topology of sparse networks from M < N perturbations. We discuss as well some aspects important for applications, such as the topology reconstruction quality and error sources, advantages and disadvantages of the suggested method, and the influence of (control) perturbations, inhomegenity, sparsity, coupling functions, and measurement noise. Some examples of networks with Chua's oscillators are presented to illustrate the reliability of the suggested technique.     

Linear fuzzy gene network models obtained from microarray data by exhaustive search

Background  

Recent technological advances in high-throughput data collection allow for experimental study of increasingly complex systems on the scale of the whole cellular genome and proteome. Gene network models are needed to interpret the resulting large and complex data sets. Rationally designed perturbations (e.g., gene knock-outs) can be used to iteratively refine hypothetical models, suggesting an approach for high-throughput biological system analysis. We introduce an approach to gene network modeling based on a scalable linear variant of fuzzy logic: a framework with greater resolution than Boolean logic models, but which, while still semi-quantitative, does not require the precise parameter measurement needed for chemical kinetics-based modeling.     

The search for common anthropogenic impacts on biodiversity: a global network

We introduce an initiative to assess and compare landscape changes related to human activities on a global scale, using a single group of invertebrates. The GLOBENET programme uses common field methodology (pitfall trapping), to appraise assemblages of ground beetles (Coleoptera, Carabidae) in visually-similar land-mosaics (urban-rural gradients). Carabids were selected as the focal taxon as they are sufficiently varied (both taxonomically and ecologically), abundant and sensitive to the environment. However, work on other taxa is comparable with the GLOBENET framework. The continuum of decreasing human pressure from city centres into the surrounding countryside was selected to represent human-caused disturbance for this initial stage of GLOBENET because these gradients can be found virtually all over the world. Through the broad-scale assessment envisioned in the GLOBENET programme, we seek to separate general, repeated effects on biodiversity from those that depend on local environments or particular biotic assemblages. Based on this understanding we aim to develop simple tools and protocols for assessing ecological effects of human-caused landscape changes, which could help to sustainably manage landscapes for biodiversity and for human requirements. For instance, the response of different functional groups of carabids to these landscape changes may help guide management practices. Further GLOBENET developments and information are available at our website: http://www.helsinki.fi/science/globenet/     

Synthesis, structure and properties of a new layered gadolinium benzenedicarboxylate with piperazine

A hydrothermal reaction of a mixture of Gd(NO₃)₃, 1,2-benzenedicarboxylic acid (1,2-BDC), piperazine, NaOH and water at 180 °C for three days under autogeneous pressure gave rise to a new compound of the formula [C₄N₂H₁₂][Gd₂(H₂O)₂(C₆H₄(COO)₂)₂] (I). The connectivity between GdO₈ distorted dodecahedra and 1,2-BDC units gives rise to a two-dimensional structure with large apertures. The fully protonated piperazine molecule occupies the middle of these apertures. The compound has favorable CH?π interactions between the benzene rings of adjacent layers and shows photoluminescence at room temperature. Crystal data: monoclinic, space group = P2₁/c (No. 14), a = 13.1671(3) Å, b = 13.7336(3) Å, c = 11.3100(1) Å, β = 115.411(1)°, v = 1847.34(6) Å³, Z = 4, R₁ = 0.0238 for 2658 reflections [I > 2σ(I)].     

Bayesian network and nonparametric heteroscedastic regression for nonlinear modeling of genetic network

We propose a new statistical method for constructing a genetic network from microarray gene expression data by using a Bayesian network. An essential point of Bayesian network construction is the estimation of the conditional distribution of each random variable. We consider fitting nonparametric regression models with heterogeneous error variances to the microarray gene expression data to capture the nonlinear structures between genes. Selecting the optimal graph, which gives the best representation of the system among genes, is still a problem to be solved. We theoretically derive a new graph selection criterion from Bayes approach in general situations. The proposed method includes previous methods based on Bayesian networks. We demonstrate the effectiveness of the proposed method through the analysis of Saccharomyces cerevisiae gene expression data newly obtained by disrupting 100 genes.