Training with noise is equivalent to Tikhonov regularization

It is well known that the addition of noise to the input data of a neural network during training can, in some circumstances, lead to significant improvements in generalization performance. Previous work has shown that such training with noise is equivalent to a form of regularization in which an extra term is added to the error function. However, the regularization term, which involves second derivatives of the error function, is not bounded below, and so can lead to difficulties if used directly in a learning algorithm based on error minimization. In this paper we show that, for the purposes of network training, the regularization term can be reduced to a positive definite form which involves only first derivatives of the network mapping. For a sum-of-squares error function, the regularization term belongs to the class of generalized Tikhonov regularizers. Direct minimization of the regularized error function provides a practical alternative to training with noise.

Building team-based working:a practical guide to organizational transformation

Building Team-based Working is designed for use by managers and consultants who are introducing team-based working into organizations. The book synthesizes knowledge about how to build team-based organizations, focusing particularly on the psychological and social processes that can facilitate or obstruct successful teamwork. Rather than advise managers on how to build effective teams, as most books in this area tend to do, this book instead focuses on how to build organizations structured around teams. The text is divided into six sections describing the six main stages of developing team-based working in an organization. The chapters follow a common structure. Each one opens with a summary of the aims and activities relevant to that stage and concludes with a selection of appropriate support materials and tools. These materials can also be downloaded from the CD accompanying the text. The advice given is based on evidence gathered by the authors over 20 years of practical management experience, research work in organizations, and consultancy across the public, manufacturing and service sectors.

Practical methods of tracking of nonstationary time series applied to real-world data

In this paper, we discuss some practical implications for implementing adaptable network algorithms applied to non-stationary time series problems. Two real world data sets, containing electricity load demands and foreign exchange market prices, are used to test several different methods, ranging from linear models with fixed parameters, to non-linear models which adapt both parameters and model order on-line. Training with the extended Kalman filter, we demonstrate that the dynamic model-order increment procedure of the resource allocating RBF network (RAN) is highly sensitive to the parameters of the novelty criterion. We investigate the use of system noise for increasing the plasticity of the Kalman filter training algorithm, and discuss the consequences for on-line model order selection. The results of our experiments show that there are advantages to be gained in tracking real world non-stationary data through the use of more complex adaptive models.

Standardisation and benchmarking for improving translator training

The worldwide growth of the translation industry requires qualified professional translators. During the last decade, we have seen an enormous increase in translator training programmes offered by universities, mainly at postgraduate level. A challenge for such university programmes is to make sure that they prepare graduates who are qualified for the needs of the diverse profession in the rapidly changing market. This means that programmes need to be developed with the market needs in mind and that they need to ensure a good match between graduates' competences and employers' requirements. This paper addresses the following questions: How can universities adapt translator training programmes to the rapidly changing industry and the accompanying changes in professional profiles? How can we reconcile the requests of the industry for graduates who have practical and professional skills with the requests of the universities for graduates who have in-depth academic knowledge and intellectual skills? What standards and benchmarks are in place to assure quality of translator training programmes? Some such developments in respect of benchmarking are illustrated at first for the United Kingdom, followed by information on the European Master’s in Translation (EMT) project, an initiative at the European level. Finally, the paper reflects on the challenges which the EMT translator competence profile poses for university programmes.

A study into the effectiveness of quality management training:a focus on tools and critical success factors

Purpose – The purpose of this paper is to investigate the effectiveness of quality management training by reviewing commonly used critical success factors and tools rather than the overall methodological approach. Design/methodology/approach – The methodology used a web-based questionnaire. It consisted of 238 questions covering 77 tools and 30 critical success factors selected from leading academic and practitioner sources. The survey had 79 usable responses and the data were analysed using relevant statistical quality management tools. The results were validated in a series of structured workshops with quality management experts. Findings – Findings show that in general most of the critical success factor statements for quality management are agreed with, although not all are implemented well. The findings also show that many quality tools are not known or understood well; and that training has an important role in raising their awareness and making sure they are used correctly. Research limitations/implications – Generalisations are limited by the UK-centric nature of the sample. Practical implications – The practical implications are discussed for organisations implementing quality management initiatives, training organisations revising their quality management syllabi and academic institutions teaching quality management. Originality/value – Most recent surveys have been aimed at methodological levels (i.e. “lean”, “Six Sigma”, “total quality management” etc.); this research proposes that this has limited value as many of the tools and critical success factors are common to most of the methodologies. Therefore, quite uniquely, this research focuses on the tools and critical success factors. Additionally, other recent comparable surveys have been less comprehensive and not focused on training issues.

Training perceptions, engagement and performance:comparing work engagement and personal role engagement

The purpose of this study was to compare two engagement constructs (work engagement and personal role engagement) with regards to their relationship with training perceptions and work role performance behaviours. It was hypothesised that personal role engagement would show incremental validity above that of work engagement at predicting work role performance behaviours and be a stronger mediator of the relationships between training perceptions and such behaviours. Questionnaire data was gathered from 304 full-time working adults in the UK. As predicted, personal role engagement was found to explain additional variance above that of work engagement for task proficiency, task adaptability, and task proactivity behaviours. Moreover, personal role engagement was a stronger mediator of the relationship between training perceptions and task proficiency as well as between training perceptions and task adaptability. Both work engagement and personal role engagement mediated the relationship between training perceptions and task proactivity to a similar degree. The findings suggest that personal role engagement has better practical utility to the HRD domain than work engagement, and indicates that future research may benefit from adopting the personal role engagement construct.