Self-organizing Neural Network Model Research Articles

A method for synthesizing neural-network models under incomplete data

Problems of synthesizing neural-network models under incomplete experimental data are described. The accuracy of the models is heavily dependent on their complexity. The proposed method allows self-organizing neural-network models of near-optimal complexity. Examples of synthesizing models of flame interferometry and growth of industrial production are presented.

Application of self-organizing competitive neural network in fault diagnosis of suck rod pumping system

This paper represents a study using Artificial Neural Network (ANN) to solve dynamometer cards auto recognition problem. A self-organizing competitive neural network model is constructed to achieve automatization of fault diagnosis by automatically clustering dynamometer cards. Model is trained and tested using data acquired from Jiangsu oil fields, China. Research results show that self-organizing competitive neural network model has good classification and generalization capabilities, and hence is applicable to the automatic fault diagnosis of suck rod pumping system.

Pattern Classification and Recognition of Invertebrate Functional Groups Using Self-Organizing Neural Networks

Self-organizing neural networks can be used to mimic non-linear systems. The main objective of this study is to make pattern classification and recognition on sampling information using two self-organizing neural network models. Invertebrate functional groups sampled in the irrigated rice field were classified and recognized using one-dimensional self-organizing map and self-organizing competitive learning neural networks. Comparisons between neural network models, distance (similarity) measures, and number of neurons were conducted. The results showed that self-organizing map and self-organizing competitive learning neural network models were effective in pattern classification and recognition of sampling information. Overall the performance of one-dimensional self-organizing map neural network was better than self-organizing competitive learning neural network. The number of neurons could determine the number of classes in the classification. Different neural network models with various distance (similarity) measures yielded similar classifications. Some differences, dependent upon the specific network structure, would be found. The pattern of an unrecognized functional group was recognized with the self-organizing neural network. A relative consistent classification indicated that the following invertebrate functional groups, terrestrial blood sucker; terrestrial flyer; tourist (nonpredatory species with no known functional role other than as prey in ecosystem); gall former; collector (gather, deposit feeder); predator and parasitoid; leaf miner; idiobiont (acarine ectoparasitoid), were classified into the same group, and the following invertebrate functional groups, external plant feeder; terrestrial crawler, walker, jumper or hunter; neustonic (water surface) swimmer (semi-aquatic), were classified into another group. It was concluded that reliable conclusions could be drawn from comparisons of different neural network models that use different distance (similarity) measures. Results with the larger consistency will be more reliable.

Prediction of flow stress for carbon steels using recurrent self-organizing neuro fuzzy networks

Mechanical properties of any material are extensively influenced by the parameters such as strain, strain rate, temperature, and its composition. The characteristics of any material such as ductility, strain hardening, strength, dynamic recovery, grain growth, and recrystallization are greatly affected by the influence of various process parameters. So, it is essential to have the knowledge of the constitutive relationships that relate different process variables to flow stress of the deforming material which estimates various parameters such as load, energy, and stress in the metal forming operations. A consistent effort has been gone into developing the constitutive equations for the detailed mathematical description of the flow curves and the aforementioned parameters for years now. Soft computing tools that concern computation of an imprecise environment and model very complex systems those are based on input-output relationship have gained significant attention in recent years. The intricacies of the mathematical modeling of the mechanical properties of the material, enticed the artificial research community to take this as a challenge. One such soft computing tool neural network is applied in this field to predict the behavior accurately. In this paper, a hybrid neural network model better known as RSONFN (Recurrent Self-Organizing Neural Network Model) is applied to predict the flow stress for carbon steels. The RSONFN is having the advantages of the well-established technologies of the artificial intelligence tools such as Fuzzy logic to capture long range data sets and neural networks. The RSONFN structure is a dynamic one as the numbers of its layers as well as the numbers of nodes in each layer of the network are not predetermined. Such an attribute differentiate it from the Multilayer perceptron which is having static structure. The results obtained by this network prove its superiority over other existing tools.

A self-organizing neural network model of the development of motion direction selectivity, orientation, and ocular dominance maps and receptive fields in V1 and MT

A self-organizing neural network model of the development of motion direction selectivity, orientation, and ocular dominance maps and receptive fields in V1 and MT

Early lexical development in a self-organizing neural network

In this paper we present a self-organizing neural network model of early lexical development called DevLex. The network consists of two self-organizing maps (a growing semantic map and a growing phonological map) that are connected via associative links trained by Hebbian learning. The model captures a number of important phenomena that occur in early lexical acquisition by children, as it allows for the representation of a dynamically changing linguistic environment in language learning. In our simulations, DevLex develops topographically organized representations for linguistic categories over time, models lexical confusion as a function of word density and semantic similarity, and shows age-of-acquisition effects in the course of learning a growing lexicon. These results match up with patterns from empirical research on lexical development, and have significant implications for models of language acquisition based on self-organizing neural networks.

Self-organising associative kernel memory for multi-domain pattern classification

This paper proposes a novel self-organising associative neural network model in terms of kernel memory. The objective of this paper is not to give a sophisticated learning scheme and its rigorous mathematical accounts but rather attempt to address a paradigm shift, which could potentially answer a number of critical issues related to the current artificial neural network architectures. In the new memory model, the notion of ‘weights’ between the nodes is totally different from that as in ordinary neural network models, in which the weights simply represent the strengths of the connections between pairs of nodes in the kernel memory each realised by a kernel unit. Hence any arduous and iterative tuning of weight parameters is not involved and thereby the neural memory does not inherently suffer from any numerically-related problems. The associative memory is constructed via a simple unsupervised learning algorithm motivated from the traditional Hebbian principle. In the simulation study, both the plasticity and performance of the novel neural network architecture are discussed within the pattern classification context through single and simultaneous multi-domain classification tasks.

A self-organizing neural network model of receptive field and map development of motion direction selectivity, orientation, and ocular dominance in V1 and MT

A self-organizing neural network model of receptive field and map development of motion direction selectivity, orientation, and ocular dominance in V1 and MT

Effects of Interactive Function Forms and Refractoryperiod in a Self-Organized Critical Model Based on Neural Networks**The project supported by National Natural Science Foundation of China and the Doctoral Foundation of the Chinese Education Commission of Ministry of Education of China under Grant Nos. 60074020 and 90203008

Based on the standard self-organizing map neural network model and an integrate-and-fire mechanism, we investigate the effect of the nonlinear interactive function on the self-organized criticality in our model. Based on these we also investigate the effect of the refractoryperiod on the self-organized criticality of the system.

A hierarchical self-organizing approach for learning the patterns of motion trajectories.

The understanding and description of object behaviors is a hot topic in computer vision. Trajectory analysis is one of the basic problems in behavior understanding, and the learning of trajectory patterns that can be used to detect anomalies and predict object trajectories is an interesting and important problem in trajectory analysis. In this paper, we present a hierarchical self-organizing neural network model and its application to the learning of trajectory distribution patterns for event recognition. The distribution patterns of trajectories are learnt using a hierarchical self-organizing neural network. Using the learned patterns, we consider anomaly detection as well as object behavior prediction. Compared with the existing neural network structures that are used to learn patterns of trajectories, our network structure has smaller scale and faster learning speed, and is thus more effective. Experimental results using two different sets of data demonstrate the accuracy and speed of our hierarchical self-organizing neural network in learning the distribution patterns of object trajectories.

Effects of Interactive Function Forms in a Self-Organized Critical Model Based on Neural Networks**The project supported by National Natural Science Foundation of China under Grant Nos. 60074020 and 90203008, and the Doctoral Foundation of the Ministry of Education of China

Based on the standard self-organizing map neural network model and an integrate-and-fire mechanism, we introduce a kind of coupled map lattice system to investigate scale-invariance behavior in the activity of model neural populations. We let the parameter β, which together with α represents the interactive strength between neurons, have different function forms, and we find the function forms and their parameters are very important to our model's avalanche dynamical behaviors, especially to the emergence of different avalanche behaviors in different areas of our system.

Self-organizing neural network model of motion processing in the visual cortex during smooth pursuit

A physiologically based neural network model was constructed to study cortical motion processing during pursuit eye movements. The model consists of three layers of computational units, simulating information processing by direction selective neurons in the primary visual cortex (V1), motion selective neurons in the middle-temporal area, and pursuit selective neurons in the middle-superior-temporal (MST) area. MST units integrate visual and eye-movement related information, and their connections develop during an unsupervised training process. The resulting MST units represent a transition from retinal to real-world reference frame. By analyzing the model connectivity, mechanisms underlying the functions performed by the network are studied.

A NEURAL NETWORK MODEL FOR MINIMUM SPANNING CIRCLE: ITS CONVERGENCE, ARCHITECTURE DESIGN AND APPLICATIONS

A self-organizing neural network model that computes the smallest circle (also called minimum spanning circle) enclosing a finite set of given points was proposed by Datta.3 In the article,3 the algorithm is stated and it is demonstrated by simulation that the center of the smallest circle can be achieved with a given level of accuracy. No rigorous proof was given in support of the simulation results. In this paper, we make a rigorous analysis of the model and mathematically prove that the model converges to the desired center of the minimum spanning circle. A suitable neural network architecture is also designed for parallel implementation of the proposed model. Time complexity of the algorithm is worked out under the proposed architecture. Extension of the proposed model to higher dimensions is discussed and demonstrated with some applications.

Type of self-organized criticality model based on neural networks.

Based on the standard self-organizing map neural network model, we introduce a kind of coupled map lattice system to investigate self-organized criticality (SOC) in the activity of model neural populations. Our system is simulated by a more detailed integrate-and-fire mechanism and a kind of local perturbation driving rule; it can display SOC behavior in a certain range of system parameters, even with period boundary condition. More importantly, when the influence of synaptic plasticity is adequately considered, we can find that our system's learning process plays a promotive role in the emergence of SOC behavior.

A sequential feature selection-based neural network approach to dynamic voltage stability estimation

Two sequential self-organizing hierarchical neural network (SHNN) models are utilized to estimate power system voltage stability. The first SHNN determines whether the power system is dynamically stable or not. The second one is used for the dynamically stable system to estimate the voltage magnitudes at all PQ buses. Tests results at different loading conditions for two test systems, WSCC nine-bus system and New England 39-bus system, are reported.

Skeletonization by a topology-adaptive self-organizing neural network

A self-organizing neural network model is proposed to generate the skeleton of a pattern. The proposed neural net is topology-adaptive and has a few advantages over other self-organizing models. The model is dynamic in the sense that it grows in size over time. The model is especially designed to produce a vector skeleton of a pattern. It works on binary patterns, dot patterns and also on gray-level patterns. Thus it provides a unified approach to skeletonization. The proposed model is highly robust to noise (boundary and interior noise) as compared to existing conventional skeletonization algorithms and is invariant under arbitrary rotation. It is also efficient in medial axis representation and in data reduction.

Impulse fault diagnosis in power transformers using self-organising map and learning vector quantisation

An artificial intelligence approach is proposed to an impulse fault diagnosis problem in oil-filled power transformers. The experiment focuses on the distinction between the effects caused by faults of a different nature and the different physical location of occurrences in a transformer winding. The proposed method involves an artificial neural network-based pattern recognition technique, to recognise the frequency responses of the winding admittance of a typical high-voltage transformer under healthy and different faulty conditions of winding insulation. It attempts to establish a correlation between the nature and site of the internal insulation fault and its associated frequency response. A self-organising neural network model has been employed as the basic pattern recogniser, to discover the significant patterns and to extract the hidden information from a set of frequency response patterns obtained from an EMTP model of the transformer with artificially simulated faults. A learning vector quantisation-based classification technique has been applied to efficiently classify visually indistinguishable response patterns. The method applied to a winding model of a high-voltage transformer, with tap changer winding, exhibited high diagnostic accuracy by successful detection and discrimination of faults of a different nature and site of occurrence.

A Model-based Interpretation of Phantom Pain — Conservative Body Schema and Flexible Somatotopy—

Some patients with amputations are known to still experience presence of the amputated limb (phantom limb). Phantom limb is found to be accompanied by reorganizations of somatotopic representations at cortical and subcortical levels due to deafferentation, which also discloses presence of body schema, i.e., integrated image of body. In addition, phantom limb is often recognized with painful sensation, i.e., phantom pain, which is suggested to be caused by abnormally high activations in thalamocortical neurons in the deafferentated area. In this paper, simulations using a self-organizing neural network model were performed to clarify contribution of axonal sprouting and body schema to reorganization of cortical representation, and to explore a possible mechanism underlying abnormally high thalamocortical activation. The results showed that axonal sprouting could cause reorganization of cortical mapping with an aid of facilitatory inputs referring to the conserved body schema even after deafferentation. In addition, removing thalamic recurrent inhibitions was shown to induce high and sustained activations in thalamocortical neurons in the deafferentated area. Consequently, this study suggested that discordance between the recognized cortical representation and physiological process referring to the body schema could give rise to phantom pain. Phantom pain is a clue disclosing how body schema is constructed and maintained, which might be an important subject also for robotics dealing with an artificial body.

A multilayer self-organizing model for convex-hull computation

A self-organizing neural-network model is proposed for computation of the convex-hull of a given set of planar points. The network evolves in such a manner that it adapts itself to the hull-vertices of the convex-hull. The proposed network consists of three layers of processors. The bottom layer computes some angles which are passed to the middle layer. The middle layer is used for computation of the minimum angle (winner selection). These information are passed to the topmost layer as well as fed back to the bottom layer. The network in the topmost layer self-organizes by labeling the hull-processors in an orderly fashion so that the final convex-hull is obtained from the topmost layer. Time complexity of the proposed model is analyzed and is compared with existing models of similar nature.

Classification and Prediction of Reagents' Roles by FRAU System with Self-Organizing Neural Network Model

Abstract The classification and prediction of the roles for reagents in reactions are presented. The same dimensional representation of various reagents independent of the number of atoms was achieved by selecting representative factors by the FRAU (Field-characterization for Reaction Analysis and Understanding) system. Training of a self-organizing model considering both negative and absent data was accomplished by modifying the original counter-propagation (CP) type of Kohonen neural network to treat absent data differently from negative data. The modified CP Kohonen neural network successfully classified the reagents and produced a reagent-roles correlation model that gives good answers predicting roles of the reagents.