Hierarchical GTM: Constructing localized non-linear projection manifolds in a principled way

It has been argued that a single two-dimensional visualization plot may not be sufficient to capture all of the interesting aspects of complex data sets, and therefore a hierarchical visualization system is desirable. In this paper we extend an existing locally linear hierarchical visualization system PhiVis ¸iteBishop98a in several directions: bf(1) We allow for em non-linear projection manifolds. The basic building block is the Generative Topographic Mapping. bf(2) We introduce a general formulation of hierarchical probabilistic models consisting of local probabilistic models organized in a hierarchical tree. General training equations are derived, regardless of the position of the model in the tree. bf(3) Using tools from differential geometry we derive expressions for local directional curvatures of the projection manifold. Like PhiVis, our system is statistically principled and is built interactively in a top-down fashion using the EM algorithm. It enables the user to interactively highlight those data in the parent visualization plot which are captured by a child model. We also incorporate into our system a hierarchical, locally selective representation of magnification factors and directional curvatures of the projection manifolds. Such information is important for further refinement of the hierarchical visualization plot, as well as for controlling the amount of regularization imposed on the local models. We demonstrate the principle of the approach on a toy data set and apply our system to two more complex 12- and 19-dimensional data sets.

Hierarchical GTM: constructing localized non-linear projection manifolds in a principled way

It has been argued that a single two-dimensional visualization plot may not be sufficient to capture all of the interesting aspects of complex data sets, and therefore a hierarchical visualization system is desirable. In this paper we extend an existing locally linear hierarchical visualization system PhiVis ¸iteBishop98a in several directions: bf(1) We allow for em non-linear projection manifolds. The basic building block is the Generative Topographic Mapping (GTM). bf(2) We introduce a general formulation of hierarchical probabilistic models consisting of local probabilistic models organized in a hierarchical tree. General training equations are derived, regardless of the position of the model in the tree. bf(3) Using tools from differential geometry we derive expressions for local directional curvatures of the projection manifold. Like PhiVis, our system is statistically principled and is built interactively in a top-down fashion using the EM algorithm. It enables the user to interactively highlight those data in the ancestor visualization plots which are captured by a child model. We also incorporate into our system a hierarchical, locally selective representation of magnification factors and directional curvatures of the projection manifolds. Such information is important for further refinement of the hierarchical visualization plot, as well as for controlling the amount of regularization imposed on the local models. We demonstrate the principle of the approach on a toy data set and apply our system to two more complex 12- and 18-dimensional data sets.

Probing three-way interactions in moderated multiple regression:Development and application of a slope difference test

Researchers often use 3-way interactions in moderated multiple regression analysis to test the joint effect of 3 independent variables on a dependent variable. However, further probing of significant interaction terms varies considerably and is sometimes error prone. The authors developed a significance test for slope differences in 3-way interactions and illustrate its importance for testing psychological hypotheses. Monte Carlo simulations revealed that sample size, magnitude of the slope difference, and data reliability affected test power. Application of the test to published data yielded detection of some slope differences that were undetected by alternative probing techniques and led to changes of results and conclusions. The authors conclude by discussing the test's applicability for psychological research. Copyright 2006 by the American Psychological Association.

Knowledge management systems:finding a way with technology

Purpose - To consider the role of technology in knowledge management in organizations, both actual and desired. Design/methodology/approach - Facilitated, computer-supported group workshops were conducted with 78 people from ten different organizations. The objective of each workshop was to review the current state of knowledge management in that organization and develop an action plan for the future. Findings - Only three organizations had adopted a strongly technology-based "solution" to knowledge management problems, and these followed three substantially different routes. There was a clear emphasis on the use of general information technology tools to support knowledge management activities, rather than the use of tools specific to knowledge management. Research limitations/implications - Further research is needed to help organizations make best use of generally available software such as intranets and e-mail for knowledge management. Many issues, especially human, relate to the implementation of any technology. Participation was restricted to organizations that wished to produce an action plan for knowledge management. The findings may therefore represent only "average" organizations, not the very best practice. Practical implications - Each organization must resolve four tensions: Between the quantity and quality of information/knowledge, between centralized and decentralized organization, between head office and organizational knowledge, and between "push" and "pull" processes. Originality/value - Although it is the group rather than an individual that determines what counts as knowledge, hardly any previous studies of knowledge management have collected data in a group context.

"High commitment" strategies:it ain't what you do; it's the way that you do it

Purpose: This paper sets out to contribute to the advancement of knowledge, particularly with regard to the processes of implementation and the role of managers engaged in such high commitment strategies and work practices. Design/methodology/approach: This study is part of a research project investigating the extent to which employee involvement predicts job performance (as well as job satisfaction, wellbeing and organisational commitment) in the NHS, using both quantitative and qualitative methods. The main focus of this paper is to present evidence from four of the 20 case studies to show the barriers to implementing employee involvement as well as highlighting the techniques and practices that have proven to be most successful. Findings: Employee involvement is used successfully by management and has enabled frontline staff to contribute their knowledge to their work. Research limitations/implications: The ethical issues of confidentiality and anonymity permeated the research process throughout. Practical implications: The link between "high commitment" strategies and organisational performance is of great interest to academics and practitioners alike. One of these "high commitment" strategies, namely employee involvement, has been an important HR strategy for the NHS in the UK. Originality/value: Other organisations can learn from the findings by implementing the successful parts. © Emerald Group Publishing Limited.

It's what you do and the way that you do it:team task, team size, and innovation-related group processes

This article describes a study of the relationships between team inputs (task type and team size) and team processes in 87 cross industry Portuguese teams, some of which had high and some low requirements to innovate. Team processes were measured using the Team Climate Inventory (TCI), which focuses on clarity of and commitment to team objectives, levels of participation, support for innovation, and quality emphases. Three hypotheses were tested. The first proposed that teams carrying out tasks with a high innovation requirement would have high scores on a measure of team processes. This was supported insofar as such teams reported higher levels of participation and support for innovation. The second hypothesis proposed that large teams would have poorer team processes. This hypothesis was confirmed. The third hypothesis concerned the interaction between size and innovation. The results suggested that large teams operating under a relatively high pressure to innovate have poorer team processes than large teams that do not have a high requirement to innovate.

Constructing localized non-linear projection manifolds in a principled way:hierarchical generative topographic mapping

It has been argued that a single two-dimensional visualization plot may not be sufficient to capture all of the interesting aspects of complex data sets, and therefore a hierarchical visualization system is desirable. In this paper we extend an existing locally linear hierarchical visualization system PhiVis (Bishop98a) in several directions: 1. We allow for em non-linear projection manifolds. The basic building block is the Generative Topographic Mapping. 2. We introduce a general formulation of hierarchical probabilistic models consisting of local probabilistic models organized in a hierarchical tree. General training equations are derived, regardless of the position of the model in the tree. 3. Using tools from differential geometry we derive expressions for local directionalcurvatures of the projection manifold. Like PhiVis, our system is statistically principled and is built interactively in a top-down fashion using the EM algorithm. It enables the user to interactively highlight those data in the parent visualization plot which are captured by a child model.We also incorporate into our system a hierarchical, locally selective representation of magnification factors and directional curvatures of the projection manifolds. Such information is important for further refinement of the hierarchical visualization plot, as well as for controlling the amount of regularization imposed on the local models. We demonstrate the principle of the approach on a toy data set andapply our system to two more complex 12- and 19-dimensional data sets.

Drug testing: a way to weed out users?

The aim of this paper is to provide managers and Human Resource executives with the basis for making drug testing policy in their organisations by presenting a critical review of existing literature on Workplace Drug Testing (WDT) and related areas which have been structured into the key areas.The key finding is whilst WDT is becoming more and more widely used, the rationale for this in terms of organizational effectiveness and safety is far from clear. Also there are significant ethical issues associated with WDT which are not always fully considered by organisations. Similarly, a cost/benefit analysis for particular organisations may well show little reason to embark on a testing policy. As a result of our review, we recommend that practitioners take a critical view of proposals introducing WDT since in many cases there is little upside to such a policy and a largely under-researched downside. There are also wider implications for society as a whole since the issue of drug taking as a whole is clearly a matter of great importance to practically every country in the world. The workplace is not at all immune from the impact of drug taking and perhaps a knee-jerk response by managers is to attempt to exclude anyone with any sort of drug habit through the use of WDT. This type of review with a specific HR focus has not been carried out before despite several calls for a more rational approach to the area.

Two way power flow using GTO thyristors

Two-way power flow is nothing new and has been in practical use using line commutated converters for at least 50 years. With these types of converters, reversal of power flow can be achieved by increasing the firing angle of the devices beyond 90 degrees thus producing a negative DC voltage. Line commutated converters have several known disadvantages including: the direct current cannot be reversed, the power factor decreases when the firing angle increases and the harmonics are high on the line current. To tackle the above problems a forced commutated converter can be used. The power factor can be unity and the harmonics can be reduced. Many researchers have used PWM with different control techniques to serve the above purposes. In each converter arm, they used a forced commutated device with an antiparallel diode. Under the rectification mode of operation the current path is preponderantly through the diodes and under the inverter operation the current flows preponderantly through the forced commutated devices. Although their results were encouraging and gave a unity power factor with nearly sinusoidal current, the main disadvantage was that there were difficulties in controlling the power factor when the system is needed to operate at lagging or leading power factor. In this work, a new idea was introduced by connecting two GTOs antiparallel instead of a diode and a GTO. A single phase system using two GTO converters which are connected in series was built. One converter operates as a rectifier and the other converter operates as an inverter. In the case of the inversion mode and in each inverter arm one GTO is operated as a diode simply by switching it always on and the other antiparallel GTO is operated as a normal device to carry the inverter current. In case of the rectification mode, in each arm one GTO is always off and the other GTP is operated as a controlled device. The main advantage is that the system can be operated at lagging or leading power factor.

Statnote 10: the two-way analysis of variance

The two-way design has been variously described as a matched-sample F-test, a simple within-subjects ANOVA, a one-way within-groups ANOVA, a simple correlated-groups ANOVA, and a one-factor repeated measures design! This confusion of terminology is likely to lead to problems in correctly identifying this analysis within commercially available software. The essential feature of the design is that each treatment is allocated by randomization to one experimental unit within each group or block. The block may be a plot of land, a single occasion in which the experiment was performed, or a human subject. The ‘blocking’ is designed to remove an aspect of the error variation and increase the ‘power’ of the experiment. If there is no significant source of variation associated with the ‘blocking’ then there is a disadvantage to the two-way design because there is a reduction in the DF of the error term compared with a fully randomised design thus reducing the ‘power’ of the analysis.

Statnote 9: The one-way analysis of variance (random effects model)

There is an alternative model of the 1-way ANOVA called the 'random effects' model or ‘nested’ design in which the objective is not to test specific effects but to estimate the degree of variation of a particular measurement and to compare different sources of variation that influence the measurement in space and/or time. The most important statistics from a random effects model are the components of variance which estimate the variance associated with each of the sources of variation influencing a measurement. The nested design is particularly useful in preliminary experiments designed to estimate different sources of variation and in the planning of appropriate sampling strategies.

Promoting scholarship - The way forward: Learning & teaching research in a complex environment – A typology

This paper provides a critical overview into a distinctive typology of Learning and Teaching Research developed at a relatively small, research-led UK University. Based upon research into staff perceptions of the relationship between learning and teaching research and practice, the model represents an holistic approach to evidence-based learning and teaching practice in Contemporary Higher Education.

A process view of knowledge management:it ain't want you do, it's the way that you do it

Knowledge management needs to consider the three related elements of people, processes and technology. Much existing work has concentrated on either people or technology, often to the exclusion of the other two elements. Yet without thinking about process ? the way people, organisations and even technology actually do things ? any implementation of a knowledge management initiative is at best risky, and at worst doomed to failure. This paper looks at various ways in which a process view has appeared, expl icitly or implicitly, in knowledge management research and practice so far, and reflects on how more 'thinking about process' might improve knowledge management in the future. Consistent with this overall viewpoint, the issues generally centre less on wha t a process view would suggest should be done, but rather on the way that it would be implemented in practice.

The making of an ideology of the third way:the evolution of christian democracy in Germany (with special reference to 1945-49)

This study investigates the search for the third way in the history of German Christian Democracy. Today, in the United Kingdom, the 'third way' is seen as a new phenomenon, a synthesis of post-war belief in the welfare state and neo-liberal conservatism. Yet it insufficiently acknowledges that the origins of third way thinking, the marriage of social justice with free market economics, of individualism with collective responsibility, are found in the early philosophies of Catholic Social Theory and Protestant Social Ethical Teaching in Germany. This study shows that in the hundred years from the 1840s to the end of the 1940s, there were Catholic and Protestant socio-ethical thinkers and political reformists in Germany who attempted to bridge the philosophical differences between liberalism and socialism, to develop a socio-economic order based on Christian moral values. It will focus on the period 1945-1949, when the CDU was founded as the first interdenominational, Christian party. The study provides the first comprehensive account of the political debates in Christian democratic groups in the Soviet, British, French and American allied occupied zones, also giving equal attention to the contribution from the Protestant wing, alongside the more widely acknowledged role of Catholics in the birth of the CDU. It examines how Christian Democrats envisaged correcting the aberrations of German history, by uniting all social classes and Christian religions in one all-embracing Volkspartei, and transforming party politics from its earlier obsession with sectarian and ideological interests towards a more pragmatic 'third way' programme. The study argues that through the making of its ideology, the CDU modified the nation's understanding of its history, re-interpreted its traditions, and redefined the meaning and perception of established political philosophies. This reveals how the ambiguity of political terminology, and the flexible practice of 'third way' politics, were an invaluable political resource in the CDU's campaign for unity, ideological legitimisation and political power.