Organisational learning, climate change and the tragedy of the commons in China

This paper ends with a brief discussion of climate change and suggests that a practical solution would be to transfer much of the current air, sea and long-haul trucking of intercontinental freight between China and Europe (and the USA) to maglev systems. First we review the potential of Asian knowledge management and organisational learning and contrast this against Western precepts finding that there seems to be little incentive to 'look after one's fellows' in China (and perhaps across Asia) outside of tight personal guanxi networks. This is likely to be the case in the intense production regions of China where little time is allowed for 'organisational learning' by the staff and there is little incentive to initiate 'knowledge management' by senior managers. Thus the 'tragedy of the commons' will be enacted by individuals, township, and provincial leaders upwards to top ministers - no one will care for the climate or pollution, only for their own group and their wealth creation prospects. Copyright © 2011 Inderscience Enterprises Ltd.

Transdisciplinarity and reflective practice in sustainable development learning

Using various extracts from the reflective commentaries of MSc students, this article explores how transdisciplinarity and reflective practice operate in the programme. It shows how learners managed the uncertainties of sustainable development through regular critical and evaluative reflections. Students were able to apprehend the several worlds making up the sustainable development project and their own personal learning journey through the various competing, complementary and occasionally contradictory perspectives, modes of learning, sources of knowledge and information. One conceptual device facilitating this process was offering an understanding of sustainable development as constituting a ‘dialogue of values’, an approach that effectively invites students to square the metaphorical circle - i.e. broadly reconciling (ecological) sustainability with (economic) development.

Uncoordinated coexisting IEEE 802.15.4 networks for machine to machine communications

IEEE 802.15.4 standard has been proposed for low power wireless personal area networks. It can be used as an important component in machine to machine (M2M) networks for data collection, monitoring and controlling functions. With an increasing number of machine devices enabled by M2M technology and equipped with 802.15.4 radios, it is likely that multiple 802.15.4 networks may be deployed closely, for example, to collect data for smart metering at residential or enterprise areas. In such scenarios, supporting reliable communications for monitoring and controlling applications is a big challenge. The problem becomes more severe due to the potential hidden terminals when the operations of multiple 802.15.4 networks are uncoordinated. In this paper, we investigate this problem from three typical scenarios and propose an analytic model to reveal how performance of coexisting 802.15.4 networks may be affected by uncoordinated operations under these scenarios. Simulations will be used to validate the analytic model. It is observed that uncoordinated operations may lead to a significant degradation of system performance in M2M applications. With the proposed analytic model, we also investigate the performance limits of the 802.15.4 networks, and the conditions under which coordinated operations may be required to support M2M applications. © 2012 Springer Science + Business Media, LLC.

Stepping Stones to Information Literacy: a PebblePad case study

Aston University has recently made PebblePad, an e-portfolio or personal learning system, available to all students within the University. The customisable Profiles within PebblePad allow students to self-declare their skills in particular areas, attaching evidence of their skills or an action plan for improvement to each statement. Formal Information Literacy (IL) teaching within Aston University is currently limited to Library & Information Services (LIS) Information Specialists delivering a maximum of one session to each student during each level of their degree. However, many of the skills are continually developed by students during the course of their academic studies. For this project, an IL skills profile was created within PebblePad, which was then promoted to groups of staff and students to complete during the academic session 2009-10. Functionality within PebblePad allowed students to share their IL skills profile, evidence, action plans or any other items they felt were appropriate with an LIS Information Specialist who was able to add comments and offer suggestions for activities to help the student to develop further. Activities were closely related to students’ coursework where possible: suggesting a student kept a short reflective log of their information searching and evaluating process for an upcoming essay, for example. Feedback on the usefulness of the IL Profile will be sought from students through focus groups and the communication tools in PebblePad. In this way, we hope to make students more aware of their IL skills and to offer IL skills support over a longer period of time than a single session can provide. We will present preliminary conclusions about the practicalities and benefits of a self-declaration approach to developing IL skills in students at Aston University.

Modeling impact of both frame collisions and frame corruptions on IEEE 802.15.4 channel access for smart grid applications

IEEE 802.15.4 standard has been recently developed for low power wireless personal area networks. It can find many applications for smart grid, such as data collection, monitoring and control functions. The performance of 802.15.4 networks has been widely studied in the literature. However the main focus has been on the modeling throughput performance with frame collisions. In this paper we propose an analytic model which can model the impact of frame collisions as well as frame corruptions due to channel bit errors. With this model the frame length can be carefully selected to improve system performance. The analytic model can also be used to study the 802.15.4 networks with interference from other co-located networks, such as IEEE 802.11 and Bluetooth networks. © 2011 Springer-Verlag.

Curvature-driven smoothing: a learning algorithm for feed-forward networks

The performance of feed-forward neural networks in real applications can be often be improved significantly if use is made of a-priori information. For interpolation problems this prior knowledge frequently includes smoothness requirements on the network mapping, and can be imposed by the addition to the error function of suitable regularization terms. The new error function, however, now depends on the derivatives of the network mapping, and so the standard back-propagation algorithm cannot be applied. In this paper, we derive a computationally efficient learning algorithm, for a feed-forward network of arbitrary topology, which can be used to minimize the new error function. Networks having a single hidden layer, for which the learning algorithm simplifies, are treated as a special case.

Dynamics of on-line gradient descent learning for multilayer neural networks

We consider the problem of on-line gradient descent learning for general two-layer neural networks. An analytic solution is presented and used to investigate the role of the learning rate in controlling the evolution and convergence of the learning process.

On-line learning in multilayer neural networks

We present an analytic solution to the problem of on-line gradient-descent learning for two-layer neural networks with an arbitrary number of hidden units in both teacher and student networks. The technique, demonstrated here for the case of adaptive input-to-hidden weights, becomes exact as the dimensionality of the input space increases.

Adaptive back-propagation in on-line learning of multilayer networks

An adaptive back-propagation algorithm is studied and compared with gradient descent (standard back-propagation) for on-line learning in two-layer neural networks with an arbitrary number of hidden units. Within a statistical mechanics framework, both numerical studies and a rigorous analysis show that the adaptive back-propagation method results in faster training by breaking the symmetry between hidden units more efficiently and by providing faster convergence to optimal generalization than gradient descent.

Learning with noise and regularizers in multilayer neural networks

We study the effect of two types of noise, data noise and model noise, in an on-line gradient-descent learning scenario for general two-layer student network with an arbitrary number of hidden units. Training examples are randomly drawn input vectors labeled by a two-layer teacher network with an arbitrary number of hidden units. Data is then corrupted by Gaussian noise affecting either the output or the model itself. We examine the effect of both types of noise on the evolution of order parameters and the generalization error in various phases of the learning process.

Finite-size effects in on-line learning of multilayer neural networks

We complement recent advances in thermodynamic limit analyses of mean on-line gradient descent learning dynamics in multi-layer networks by calculating fluctuations possessed by finite dimensional systems. Fluctuations from the mean dynamics are largest at the onset of specialisation as student hidden unit weight vectors begin to imitate specific teacher vectors, increasing with the degree of symmetry of the initial conditions. In light of this, we include a term to stimulate asymmetry in the learning process, which typically also leads to a significant decrease in training time.

Learning with regularizers in multilayer neural networks

We study the effect of regularization in an on-line gradient-descent learning scenario for a general two-layer student network with an arbitrary number of hidden units. Training examples are randomly drawn input vectors labelled by a two-layer teacher network with an arbitrary number of hidden units which may be corrupted by Gaussian output noise. We examine the effect of weight decay regularization on the dynamical evolution of the order parameters and generalization error in various phases of the learning process, in both noiseless and noisy scenarios.

Globally optimal parameters for on-line learning in multilayer neural networks

We present a framework for calculating globally optimal parameters, within a given time frame, for on-line learning in multilayer neural networks. We demonstrate the capability of this method by computing optimal learning rates in typical learning scenarios. A similar treatment allows one to determine the relevance of related training algorithms based on modifications to the basic gradient descent rule as well as to compare different training methods.

Globally optimal on-line learning rules for multi-layer neural networks

We present a method for determining the globally optimal on-line learning rule for a soft committee machine under a statistical mechanics framework. This rule maximizes the total reduction in generalization error over the whole learning process. A simple example demonstrates that the locally optimal rule, which maximizes the rate of decrease in generalization error, may perform poorly in comparison.

The role of biases in on-line learning of two-layer networks

The influence of biases on the learning dynamics of a two-layer neural network, a normalized soft-committee machine, is studied for on-line gradient descent learning. Within a statistical mechanics framework, numerical studies show that the inclusion of adjustable biases dramatically alters the learning dynamics found previously. The symmetric phase which has often been predominant in the original model all but disappears for a non-degenerate bias task. The extended model furthermore exhibits a much richer dynamical behavior, e.g. attractive suboptimal symmetric phases even for realizable cases and noiseless data.