A novel adaptive weight selection algorithm for multi-objective multi-agent reinforcement learning

To solve multi-objective problems, multiple reward signals are often scalarized into a single value and further processed using established single-objective problem solving techniques. While the field of multi-objective optimization has made many advances in applying scalarization techniques to obtain good solution trade-offs, the utility of applying these techniques in the multi-objective multi-agent learning domain has not yet been thoroughly investigated. Agents learn the value of their decisions by linearly scalarizing their reward signals at the local level, while acceptable system wide behaviour results. However, the non-linear relationship between weighting parameters of the scalarization function and the learned policy makes the discovery of system wide trade-offs time consuming. Our first contribution is a thorough analysis of well known scalarization schemes within the multi-objective multi-agent reinforcement learning setup. The analysed approaches intelligently explore the weight-space in order to find a wider range of system trade-offs. In our second contribution, we propose a novel adaptive weight algorithm which interacts with the underlying local multi-objective solvers and allows for a better coverage of the Pareto front. Our third contribution is the experimental validation of our approach by learning bi-objective policies in self-organising smart camera networks. We note that our algorithm (i) explores the objective space faster on many problem instances, (ii) obtained solutions that exhibit a larger hypervolume, while (iii) acquiring a greater spread in the objective space.

An alternating method for Cauchy problems for Helmholtz-type operators in non-homogeneous medium

Kozlov & Maz'ya (1989, Algebra Anal., 1, 144–170) proposed an alternating iterative method for solving Cauchy problems for general strongly elliptic and formally self-adjoint systems. However, in many applied problems, operators appear that do not satisfy these requirements, e.g. Helmholtz-type operators. Therefore, in this study, an alternating procedure for solving Cauchy problems for self-adjoint non-coercive elliptic operators of second order is presented. A convergence proof of this procedure is given.

Guiding local regression using visualisation

Solving many scientific problems requires effective regression and/or classification models for large high-dimensional datasets. Experts from these problem domains (e.g. biologists, chemists, financial analysts) have insights into the domain which can be helpful in developing powerful models but they need a modelling framework that helps them to use these insights. Data visualisation is an effective technique for presenting data and requiring feedback from the experts. A single global regression model can rarely capture the full behavioural variability of a huge multi-dimensional dataset. Instead, local regression models, each focused on a separate area of input space, often work better since the behaviour of different areas may vary. Classical local models such as Mixture of Experts segment the input space automatically, which is not always effective and it also lacks involvement of the domain experts to guide a meaningful segmentation of the input space. In this paper we addresses this issue by allowing domain experts to interactively segment the input space using data visualisation. The segmentation output obtained is then further used to develop effective local regression models.

The role of attributions in reactions to job relocation

Job relocation refers to the process of simultaneously moving to a new job and house and this can cause considerable stress for the relocator and his/her family. Based upon an attributional analysis, we predicted that negative psychological reactions would be a function of (1) number of relocation problems, and (2) making pessimistic attributions for relocation problems (that is, the tendency to attribute negative events to internal, stable and global causes). Furthermore, these factors should interact, such that individuals with many relocation problems who also make pessimistic attributions will experience the worst psychological reactions. The results from a cross-sectional survey of 93 relocators supported these predictions. As expected, those relocators who had many relocation problems and made pessimistic attributions reported the worst mental health and relocation-specific stress. In addition, a reanalysis of a longitudinal study of relocators by Martin (1996) also supported the above predictions using attributions of perceived control. Furthermore, the relocators predicted to be most at risk (many problems}/low control) reported the worst changes in mental health during the course of the move.

Corruption in the post-Soviet workplace: the experiences of recent graduates in contemporary Ukraine

While Ukraine was bestowed market economy status by the European Union in 2005 its labour market still endures many structural problems. By exploring the experiences of young graduate employees this article highlights the difficultly in obtaining work within Ukraine's labour market and the problems they face once they have secured employment. Rather than seeing the development of a transparent labour market the collapse of the command economy has seen a relatively closed system develop. The article demonstrates how many jobs are secured through the use of connections or the demanding, and payment, of bribes.The situation does not improve once graduates obtain long-term employment. Interviewees discuss the lack of job security, the informal payment of wages and the lack of legal protection from corrupt employer practices. The article has broader resonance outside of the Ukrainian case study as the discussion of workplace corruption highlights how the issue is concerned with much more than simply cash based transactions and how those that endure it are likely to turn to the informal economy for employment.

Guiding local regression using visualisation

Solving many scientific problems requires effective regression and/or classification models for large high-dimensional datasets. Experts from these problem domains (e.g. biologists, chemists, financial analysts) have insights into the domain which can be helpful in developing powerful models but they need a modelling framework that helps them to use these insights. Data visualisation is an effective technique for presenting data and requiring feedback from the experts. A single global regression model can rarely capture the full behavioural variability of a huge multi-dimensional dataset. Instead, local regression models, each focused on a separate area of input space, often work better since the behaviour of different areas may vary. Classical local models such as Mixture of Experts segment the input space automatically, which is not always effective and it also lacks involvement of the domain experts to guide a meaningful segmentation of the input space. In this paper we addresses this issue by allowing domain experts to interactively segment the input space using data visualisation. The segmentation output obtained is then further used to develop effective local regression models.

N-dimensional electron in an anharmonic potential:the large-N limit

We show that the energy levels predicted by a frac(1, N)-expansion method for an N-dimensional electron in an anharmonic potential are always lower than the exact energy levels but monotonically converge toward their exact eigenstates for higher ordered corrections. The technique allows a systematic approach for quantum many body problems in a confined potential and explains the remarkable agreement of such approximate theories when compared with numerical results.

Making a difference:critical linguistic analysis in a legal context

One of the major problems for Critical Discourse Analysts is how to move on from their insightful critical analyses to successfully 'acting on the world in order to transform it'. This paper discusses, with detailed exemplification, some of the areas where linguists have moved beyond description to acting on and changing the world. Examples from three murder trials show how essential it is, in order to protect the rights of witnesses and defendants, to have audio records of significant interviews with police officers. The article moves on to discuss the potentially serious consequences of the many communicative problems inherent in legal/lay interaction and illustrates a few of the linguist-led improvements to important texts. Finally, the article turns to the problems of using linguistic data to try to determine the geographical origin of asylum seekers. The intention of the article is to act as a call to arms to linguists; it concludes with the observation that 'innumerable mountains remain for those with a critical linguistic perspective who would like to try to move one'. © 2011 John Benjamins Publishing Company.

Parallel Problem Solving from Nature – PPSN XIV:14th International Conference, Edinburgh, UK, September 17-21, 2016, Proceedings

This book constitutes the refereed proceedings of the 14th International Conference on Parallel Problem Solving from Nature, PPSN 2016, held in Edinburgh, UK, in September 2016. The total of 93 revised full papers were carefully reviewed and selected from 224 submissions. The meeting began with four workshops which offered an ideal opportunity to explore specific topics in intelligent transportation Workshop, landscape-aware heuristic search, natural computing in scheduling and timetabling, and advances in multi-modal optimization. PPSN XIV also included sixteen free tutorials to give us all the opportunity to learn about new aspects: gray box optimization in theory; theory of evolutionary computation; graph-based and cartesian genetic programming; theory of parallel evolutionary algorithms; promoting diversity in evolutionary optimization: why and how; evolutionary multi-objective optimization; intelligent systems for smart cities; advances on multi-modal optimization; evolutionary computation in cryptography; evolutionary robotics - a practical guide to experiment with real hardware; evolutionary algorithms and hyper-heuristics; a bridge between optimization over manifolds and evolutionary computation; implementing evolutionary algorithms in the cloud; the attainment function approach to performance evaluation in EMO; runtime analysis of evolutionary algorithms: basic introduction; meta-model assisted (evolutionary) optimization. The papers are organized in topical sections on adaption, self-adaption and parameter tuning; differential evolution and swarm intelligence; dynamic, uncertain and constrained environments; genetic programming; multi-objective, many-objective and multi-level optimization; parallel algorithms and hardware issues; real-word applications and modeling; theory; diversity and landscape analysis.

Distributed optimization in transportation and logistics networks

Many important problems in communication networks, transportation networks, and logistics networks are solved by the minimization of cost functions. In general, these can be complex optimization problems involving many variables. However, physicists noted that in a network, a node variable (such as the amount of resources of the nodes) is connected to a set of link variables (such as the flow connecting the node), and similarly each link variable is connected to a number of (usually two) node variables. This enables one to break the problem into local components, often arriving at distributive algorithms to solve the problems. Compared with centralized algorithms, distributed algorithms have the advantages of lower computational complexity, and lower communication overhead. Since they have a faster response to local changes of the environment, they are especially useful for networks with evolving conditions. This review will cover message-passing algorithms in applications such as resource allocation, transportation networks, facility location, traffic routing, and stability of power grids.

What to do when K-means clustering fails:a simple yet principled alternative algorithm

The K-means algorithm is one of the most popular clustering algorithms in current use as it is relatively fast yet simple to understand and deploy in practice. Nevertheless, its use entails certain restrictive assumptions about the data, the negative consequences of which are not always immediately apparent, as we demonstrate. While more flexible algorithms have been developed, their widespread use has been hindered by their computational and technical complexity. Motivated by these considerations, we present a flexible alternative to K-means that relaxes most of the assumptions, whilst remaining almost as fast and simple. This novel algorithm which we call MAP-DP (maximum a-posteriori Dirichlet process mixtures), is statistically rigorous as it is based on nonparametric Bayesian Dirichlet process mixture modeling. This approach allows us to overcome most of the limitations imposed by K-means. The number of clusters K is estimated from the data instead of being fixed a-priori as in K-means. In addition, while K-means is restricted to continuous data, the MAP-DP framework can be applied to many kinds of data, for example, binary, count or ordinal data. Also, it can efficiently separate outliers from the data. This additional flexibility does not incur a significant computational overhead compared to K-means with MAP-DP convergence typically achieved in the order of seconds for many practical problems. Finally, in contrast to K-means, since the algorithm is based on an underlying statistical model, the MAP-DP framework can deal with missing data and enables model testing such as cross validation in a principled way. We demonstrate the simplicity and effectiveness of this algorithm on the health informatics problem of clinical sub-typing in a cluster of diseases known as parkinsonism.

Odour pollution problems within the district of the Wrekin council

This thesis is concerned with various aspects of Air Pollution due to smell, the impact it has on communities exposed to it, the means by which it may be controlled and the manner in which a local authority may investigate the problems it causes. The approach is a practical one drawing on examples occurring within a Local Authority's experience and for that reason the research is anecdotal and is not a comprehensive treatise on the full range of options available. Odour Pollution is not yet a well organised discipline and might be considered esoteric as it is necessary to incorporate elements of science and the humanities. It has been necessary to range widely across a number of aspects of the subject so that discussion is often restricted but many references have been included to enable a reader to pursue a particular point in greater depth. In a `fuzzy' subject there is often a yawning gap separating theory and practice, thus case studies have been used to illustrate the interplay of various disciplines in resolution of a problem. The essence of any science is observation and measurement. Observation has been made of the spread of odour pollution through a community and also of relevant meterological data so that a mathematical model could be constructed and its predictions checked. It has been used to explore the results of some options for odour control. Measurements of odour perception and human behaviour seldom have the precision and accuracy of the physical sciences. However methods of social research enabled individual perception of odour pollution to be quantified and an insight gained into reaction of a community exposed to it. Odours have four attributes that can be measured and together provide a complete description of its perception. No objective techniques of measurement have yet been developed but in this thesis simple, structured procedures of subjective assessment have been improvised and their use enabled the functioning of the components of an odour control system to be assessed. Such data enabled the action of the system to be communicated using terms that are understood by a non specialist audience.

Design of a genetic algorithm for bi-objective flow shop scheduling problems with re-entrant jobs

This paper presents a simulated genetic algorithm (GA) model of scheduling the flow shop problem with re-entrant jobs. The objective of this research is to minimize the weighted tardiness and makespan. The proposed model considers that the jobs with non-identical due dates are processed on the machines in the same order. Furthermore, the re-entrant jobs are stochastic as only some jobs are required to reenter to the flow shop. The tardiness weight is adjusted once the jobs reenter to the shop. The performance of the proposed GA model is verified by a number of numerical experiments where the data come from the case company. The results show the proposed method has a higher order satisfaction rate than the current industrial practices.

Objective clinical performance of ‘comfort-enhanced' daily disposable soft contact lenses

Purpose: To examine the objective clinical performance of ‘comfort-enhanced’ daily disposable contact lenses over a 16-h day. Methods: Four contact lenses (Hilafilcon B, Etafilcon A Plus, Nelfilcon A and Nelfilcon A Plus) were evaluated in an investigator masked, open label trial at the end of a week’s bilateral wear. Pre-lens noninvasive tear break-up time (PL-NITBUT), tear prism height, bulbar hyperaemia and ocular surface temperature (OST) were measured with the lens in situ at 8, 12 and 16 h of wear. Results: There was no difference between how many hours the lenses types were worn each day (F = 0.90, p = 0.44). The PL-NITBUT decreased with the duration of daily lens wear (F = 32.0, p < 0.001) and was more stable with Nelfilcon A Plus (F = 6.00, p = 0.002) than with the other lenses evaluated. Bulbar blood vessels increased in coverage (F = 11.5, p < 0.001) but not overall redness (F = 0.0, p = 0.99) with the duration of daily lens wear, but there was no difference between the lenses (p > 0.05). The tear prism height decreased with the duration of daily wear (F = 27.0, p < 0.001) and differed between lenses (F = 2.9, p = 0.04). The OST decreased with the duration of lens wear (F = 119.7, p < 0.001) and was reduced by daily disposable lens wear (F = 7.88, p < 0.001), but did not differ between lenses (F = 0.88, p = 0.45). Conclusions: Objective measures of tear film indicated a difference between the lenses evaluated for PLNITBUT and tear prism height, but not for wearing time or bulbar conjunctival hyperaemia. Therefore clinical benefits of daily disposable ‘comfort enhancing’ contact lenses can be measured, but challenges remain in producing contact lenses that do not compromise anterior eye physiology over the whole day. 2010 British Contact Lens Association. Published by Elsevier Ltd. All rights reserved.

Editorial

This second issue of Knowledge Management Research & Practice (KMRP) continues the international nature of the first issue, with papers from authors based on four different continents. There are five regular papers, plus the first of what is intended to be an occasional series of 'position papers' from respected figures in the knowledge management field, who have specific issues they wish to raise from a personal standpoint. The first two regular papers are both based on case studies. The first is 'Aggressively pursuing knowledge management over two years: a case study a US government organization' by Jay Liebowitz. Liebowitz is well known to both academics and practictioners as an author on knowledge management and knowledge based systems. Government departments in many Western countries must soon face up to the problems that will occur as the 'baby boomer' generation reaches retirement age over the next decade. This paper describes how one particular US government organization has attempted to address this situation (and others) through the introduction of a knowledge management initiative. The second case study paper is 'Knowledge creation through the synthesizing capability of networked strategic communities: case study on new product development in Japan' by Mitsuru Kodama. This paper looks at the importance of strategic communities - communities that have strategic relevance and support - in knowledge management. Here, the case study organization is Nippon Telegraph and Telephone Corporation (NTT), a Japanese telecommunication firm. The third paper is 'Knowledge management and intellectual capital: an empirical examination of current practice in Australia' by Albert Zhou and Dieter Fink. This paper reports the results of a survey carried out in 2001, exploring the practices relating to knowledge management and intellectual capital in Australia and the relationship between them. The remaining two regular papers are conceptual in nature. The fourth is 'The enterprise knowledge dictionary' by Stuart Galup, Ronald Dattero and Richard Hicks. Galup, Dattero and Hicks propose the concept of an enterprise knowledge dictionary and its associated knowledge management system architecture as offering the appropriate form of information technology to support various different types of knowledge sources, while behaving as a single source from the user's viewpoint. The fifth and final regular paper is 'Community of practice and metacapabilities' by Geri Furlong and Leslie Johnson. This paper looks at the role of communities of practice in learning in organizations. Its emphasis is on metacapabilities - the properties required to learn, develop and apply skills. This discussion takes work on learning and core competences to a higher level. Finally, this issue includes a position paper 'Innovation as an objective of knowledge management. Part I: the landscape of management' by Dave Snowden. Snowden has been highly visible in the knowledge management community thanks to his role as the Director of IBM Global Services' Canolfan Cynefin Centre. He has helped many government and private sector organizations to consider their knowledge management problems and strategies. This, the first of two-part paper, is inspired by the notion of complexity. In it, Snowden calls for what he sees as a 20th century emphasis on designed systems for knowledge management to be consigned to history, and replaced by a 21st century emphasis on emergence. Letters to the editor on this, or any other topic related to knowledge management research and practice, are welcome. We trust that you will find the contributions stimulating, and again invite you to contribute your own paper(s) to future issues of KMRP.