Self-organising neural networks for adaptive control

Self-organizing neural networks have been implemented in a wide range of application areas such as speech processing, image processing, optimization and robotics. Recent variations to the basic model proposed by the authors enable it to order state space using a subset of the input vector and to apply a local adaptation procedure that does not rely on a predefined test duration limit. Both these variations have been incorporated into a new feature map architecture that forms an integral part of an Hybrid Learning System (HLS) based on a genetic-based classifier system. Problems are represented within HLS as objects characterized by environmental features. Objects controlled by the system have preset targets set against a subset of their features. The system's objective is to achieve these targets by evolving a behavioural repertoire that efficiently explores and exploits the problem environment. Feature maps encode two types of knowledge within HLS — long-term memory traces of useful regularities within the environment and the classifier performance data calibrated against an object's feature states and targets. Self-organization of these networks constitutes non-genetic-based (experience-driven) learning within HLS. This paper presents a description of the HLS architecture and an analysis of the modified feature map implementing associative memory. Initial results are presented that demonstrate the behaviour of the system on a simple control task.

Neural networks for power systems alarm handling

A multi-layered architecture of self-organizing neural networks is being developed as part of an intelligent alarm processor to analyse a stream of power grid fault messages and provide a suggested diagnosis of the fault location. Feedback concerning the accuracy of the diagnosis is provided by an object-oriented grid simulator which acts as an external supervisor to the learning system. The utilization of artificial neural networks within this environment should result in a powerful generic alarm processor which will not require extensive training by a human expert to produce accurate results.

Adaptive alarm processor for fault diagnosis on power transmission networks

The authors describe a learning classifier system (LCS) which employs genetic algorithms (GA) for adaptive online diagnosis of power transmission network faults. The system monitors switchgear indications produced by a transmission network, reporting fault diagnoses on any patterns indicative of faulted components. The system evaluates the accuracy of diagnoses via a fault simulator developed by National Grid Co. and adapts to reflect the current network topology by use of genetic algorithms.

Neurofuzzy mixture of experts network parallel learning and model construction algorithms

A connection between a fuzzy neural network model with the mixture of experts network (MEN) modelling approach is established. Based on this linkage, two new neuro-fuzzy MEN construction algorithms are proposed to overcome the curse of dimensionality that is inherent in the majority of associative memory networks and/or other rule based systems. The first construction algorithm employs a function selection manager module in an MEN system. The second construction algorithm is based on a new parallel learning algorithm in which each model rule is trained independently, for which the parameter convergence property of the new learning method is established. As with the first approach, an expert selection criterion is utilised in this algorithm. These two construction methods are equivalent in their effectiveness in overcoming the curse of dimensionality by reducing the dimensionality of the regression vector, but the latter has the additional computational advantage of parallel processing. The proposed algorithms are analysed for effectiveness followed by numerical examples to illustrate their efficacy for some difficult data based modelling problems.

Dual-orthogonal radial basis function networks for nonlinear time series prediction

A new structure of Radial Basis Function (RBF) neural network called the Dual-orthogonal RBF Network (DRBF) is introduced for nonlinear time series prediction. The hidden nodes of a conventional RBF network compare the Euclidean distance between the network input vector and the centres, and the node responses are radially symmetrical. But in time series prediction where the system input vectors are lagged system outputs, which are usually highly correlated, the Euclidean distance measure may not be appropriate. The DRBF network modifies the distance metric by introducing a classification function which is based on the estimation data set. Training the DRBF networks consists of two stages. Learning the classification related basis functions and the important input nodes, followed by selecting the regressors and learning the weights of the hidden nodes. In both cases, a forward Orthogonal Least Squares (OLS) selection procedure is applied, initially to select the important input nodes and then to select the important centres. Simulation results of single-step and multi-step ahead predictions over a test data set are included to demonstrate the effectiveness of the new approach.

Resilient responses to climate change: exploring organisational learning in networked development organisations

Drawing from the organisational learning and governance literature, this paper assesses four internationally networked governmental and non‐governmental organisations in the UK addressing climate change. We analyse how those concerned understand the climate change crisis, what mechanisms are put in place to address information flows, and what evidence there is of learning through sharing information between the organisational headquarters and their regional offices. The most striking finding is the evidence of learning that largely depends on ad‐hoc informal processes and shadow networks.

Incorporating creativity into teachers practice and self-concept of professional identity

The study explores what happens to teachers practice and ’ professional identity when they adopt a collaborative action research approach to teaching and involve external creative partners and a university mentor. The teachers aim to nurture and develop the creative potential of their learners through empowering them to make decisions for themselves about their own progress and learning directions. The teachers worked creatively and collaboratively designing creative teaching and learning methods in support of pupils with language and communication difficulties. The respondents are from an English special school, primary school and girls secondary school. A mixed methods methodology is adopted. Gains in teacher confidence and capability were identified in addition to shifts in values that impacted directly on their self-concept of what it is to be an effective teacher promoting effective learning. The development of their professional identities within a team ethos included them being able to make decisions about learning that are based on the educational potential of learners that they proved resulted in elevated standards achieved by this group of learners. They were able to justify their actions on established educational principles. Tensions however were revealed between what they perceived as their normal required professionalism imposed by external agencies and the enhanced professionalism experienced working through the project where they were able to integrate theory and practice.

An overview of the use of neural networks for data mining tasks

In the recent years, the area of data mining has been experiencing considerable demand for technologies that extract knowledge from large and complex data sources. There has been substantial commercial interest as well as active research in the area that aim to develop new and improved approaches for extracting information, relationships, and patterns from large datasets. Artificial neural networks (NNs) are popular biologically-inspired intelligent methodologies, whose classification, prediction, and pattern recognition capabilities have been utilized successfully in many areas, including science, engineering, medicine, business, banking, telecommunication, and many other fields. This paper highlights from a data mining perspective the implementation of NN, using supervised and unsupervised learning, for pattern recognition, classification, prediction, and cluster analysis, and focuses the discussion on their usage in bioinformatics and financial data analysis tasks. © 2012 Wiley Periodicals, Inc.

When long-range zero-lag synchronization is feasible in cortical networks

Many studies have reported long-range synchronization of neuronal activity between brain areas, in particular in the beta and gamma bands with frequencies in the range of 14–30 and 40–80 Hz, respectively. Several studies have reported synchrony with zero phase lag, which is remarkable considering the synaptic and conduction delays inherent in the connections between distant brain areas. This result has led to many speculations about the possible functional role of zero-lag synchrony, such as for neuronal communication, attention, memory, and feature binding. However, recent studies using recordings of single-unit activity and local field potentials report that neuronal synchronization may occur with non-zero phase lags. This raises the questions whether zero-lag synchrony can occur in the brain and, if so, under which conditions. We used analytical methods and computer simulations to investigate which connectivity between neuronal populations allows or prohibits zero-lag synchrony. We did so for a model where two oscillators interact via a relay oscillator. Analytical results and computer simulations were obtained for both type I Mirollo–Strogatz neurons and type II Hodgkin–Huxley neurons. We have investigated the dynamics of the model for various types of synaptic coupling and importantly considered the potential impact of Spike-Timing Dependent Plasticity (STDP) and its learning window. We confirm previous results that zero-lag synchrony can be achieved in this configuration. This is much easier to achieve with Hodgkin–Huxley neurons, which have a biphasic phase response curve, than for type I neurons. STDP facilitates zero-lag synchrony as it adjusts the synaptic strengths such that zero-lag synchrony is feasible for a much larger range of parameters than without STDP.

Complex-valued B-spline neural networks for modeling and inverting Hammerstein systems

Many communication signal processing applications involve modelling and inverting complex-valued (CV) Hammerstein systems. We develops a new CV B-spline neural network approach for efficient identification of the CV Hammerstein system and effective inversion of the estimated CV Hammerstein model. Specifically, the CV nonlinear static function in the Hammerstein system is represented using the tensor product from two univariate B-spline neural networks. An efficient alternating least squares estimation method is adopted for identifying the CV linear dynamic model’s coefficients and the CV B-spline neural network’s weights, which yields the closed-form solutions for both the linear dynamic model’s coefficients and the B-spline neural network’s weights, and this estimation process is guaranteed to converge very fast to a unique minimum solution. Furthermore, an accurate inversion of the CV Hammerstein system can readily be obtained using the estimated model. In particular, the inversion of the CV nonlinear static function in the Hammerstein system can be calculated effectively using a Gaussian-Newton algorithm, which naturally incorporates the efficient De Boor algorithm with both the B-spline curve and first order derivative recursions. The effectiveness of our approach is demonstrated using the application to equalisation of Hammerstein channels.

Performing knowledge requirements analysis for public organisations in a Virtual Learning Environment: a social network analysis approach

This paper describes an application of Social Network Analysis methods for identification of knowledge demands in public organisations. Affiliation networks established in a postgraduate programme were analysed. The course was executed in a distance education mode and its students worked on public agencies. Relations established among course participants were mediated through a virtual learning environment using Moodle. Data available in Moodle may be extracted using knowledge discovery in databases techniques. Potential degrees of closeness existing among different organisations and among researched subjects were assessed. This suggests how organisations could cooperate for knowledge management and also how to identify their common interests. The study points out that closeness among organisations and research topics may be assessed through affiliation networks. This opens up opportunities for applying knowledge management between organisations and creating communities of practice. Concepts of knowledge management and social network analysis provide the theoretical and methodological basis.

Joint multiple dictionary learning for tensor sparse coding

Traditional dictionary learning algorithms are used for finding a sparse representation on high dimensional data by transforming samples into a one-dimensional (1D) vector. This 1D model loses the inherent spatial structure property of data. An alternative solution is to employ Tensor Decomposition for dictionary learning on their original structural form —a tensor— by learning multiple dictionaries along each mode and the corresponding sparse representation in respect to the Kronecker product of these dictionaries. To learn tensor dictionaries along each mode, all the existing methods update each dictionary iteratively in an alternating manner. Because atoms from each mode dictionary jointly make contributions to the sparsity of tensor, existing works ignore atoms correlations between different mode dictionaries by treating each mode dictionary independently. In this paper, we propose a joint multiple dictionary learning method for tensor sparse coding, which explores atom correlations for sparse representation and updates multiple atoms from each mode dictionary simultaneously. In this algorithm, the Frequent-Pattern Tree (FP-tree) mining algorithm is employed to exploit frequent atom patterns in the sparse representation. Inspired by the idea of K-SVD, we develop a new dictionary update method that jointly updates elements in each pattern. Experimental results demonstrate our method outperforms other tensor based dictionary learning algorithms.

Training on the use of a bespoke continuing professional development framework improves the quality of CPD records

BACKGROUND: Using continuing professional development (CPD) as part of the revalidation of pharmacy professionals has been proposed in the UK but not implemented. We developed a CPD Outcomes Framework (‘the framework’) for scoring CPD records, where the score range was -100 to +150 based on demonstrable relevance and impact of the CPD on practice. OBJECTIVE: This exploratory study aimed to test the outcome of training people to use the framework, through distance-learning material (active intervention), by comparing CPD scores before and after training. SETTING: Pharmacy professionals were recruited in the UK in Reading, Banbury, Southampton, Kingston-upon-Thames and Guildford in 2009. METHOD: We conducted a randomised, double-blinded, parallel-group, before and after study. The control group simply received information on new CPD requirements through the post; the active intervention group also received the framework and associated training. Altogether 48 participants (25 control, 23 active) completed the study. All participants submitted CPD records to the research team before and after receiving the posted resources. The records (n=226) were scored blindly by the researchers using the framework. A subgroup of CPD records (n=96) submitted first (before-stage) and rewritten (after-stage) were analysed separately. MAIN OUTCOME MEASURE: Scores for CPD records received before and after distributing group-dependent material through the post. RESULTS: Using a linear-regression model both analyses found an increase in CPD scores in favour of the active intervention group. For the complete set of records, the effect was a mean difference of 9.9 (95% CI = 0.4 to 19.3), p-value = 0.04. For the subgroup of rewritten records, the effect was a mean difference of 17.3 (95% CI = 5.6 to 28.9), p-value = 0.0048. CONCLUSION: The intervention improved participants’ CPD behaviour. Training pharmacy professionals to use the framework resulted in better CPD activities and CPD records, potentially helpful for revalidation of pharmacy professionals. IMPACT: • Using a bespoke Continuing Professional Development outcomes framework improves the value of pharmacy professionals’ CPD activities and CPD records, with the potential to improve patient care. • The CPD outcomes framework could be helpful to pharmacy professionals internationally who want to improve the quality of their CPD activities and CPD records. • Regulators and officials across Europe and beyond can assess the suitability of the CPD outcomes framework for use in pharmacy CPD and revalidation in their own setting.