Nonlinear model structure design and construction using orthogonal least squares and D-optimality design

A very efficient learning algorithm for model subset selection is introduced based on a new composite cost function that simultaneously optimizes the model approximation ability and model robustness and adequacy. The derived model parameters are estimated via forward orthogonal least squares, but the model subset selection cost function includes a D-optimality design criterion that maximizes the determinant of the design matrix of the subset to ensure the model robustness, adequacy, and parsimony of the final model. The proposed approach is based on the forward orthogonal least square (OLS) algorithm, such that new D-optimality-based cost function is constructed based on the orthogonalization process to gain computational advantages and hence to maintain the inherent advantage of computational efficiency associated with the conventional forward OLS approach. Illustrative examples are included to demonstrate the effectiveness of the new approach.

A mixture of experts network structure construction algorithm for modelling and control

This paper introduces a new fast, effective and practical model structure construction algorithm for a mixture of experts network system utilising only process data. The algorithm is based on a novel forward constrained regression procedure. Given a full set of the experts as potential model bases, the structure construction algorithm, formed on the forward constrained regression procedure, selects the most significant model base one by one so as to minimise the overall system approximation error at each iteration, while the gate parameters in the mixture of experts network system are accordingly adjusted so as to satisfy the convex constraints required in the derivation of the forward constrained regression procedure. The procedure continues until a proper system model is constructed that utilises some or all of the experts. A pruning algorithm of the consequent mixture of experts network system is also derived to generate an overall parsimonious construction algorithm. Numerical examples are provided to demonstrate the effectiveness of the new algorithms. The mixture of experts network framework can be applied to a wide variety of applications ranging from multiple model controller synthesis to multi-sensor data fusion.

Neurofuzzy design and model construction of nonlinear dynamical processes from data

A common problem in many data based modelling algorithms such as associative memory networks is the problem of the curse of dimensionality. In this paper, a new two-stage neurofuzzy system design and construction algorithm (NeuDeC) for nonlinear dynamical processes is introduced to effectively tackle this problem. A new simple preprocessing method is initially derived and applied to reduce the rule base, followed by a fine model detection process based on the reduced rule set by using forward orthogonal least squares model structure detection. In both stages, new A-optimality experimental design-based criteria we used. In the preprocessing stage, a lower bound of the A-optimality design criterion is derived and applied as a subset selection metric, but in the later stage, the A-optimality design criterion is incorporated into a new composite cost function that minimises model prediction error as well as penalises the model parameter variance. The utilisation of NeuDeC leads to unbiased model parameters with low parameter variance and the additional benefit of a parsimonious model structure. Numerical examples are included to demonstrate the effectiveness of this new modelling approach for high dimensional inputs.

Nonlinear model structure detection using optimum experimental design and orthogonal least squares

A very efficient learning algorithm for model subset selection is introduced based on a new composite cost function that simultaneously optimizes the model approximation ability and model adequacy. The derived model parameters are estimated via forward orthogonal least squares, but the subset selection cost function includes an A-optimality design criterion to minimize the variance of the parameter estimates that ensures the adequacy and parsimony of the final model. An illustrative example is included to demonstrate the effectiveness of the new approach.

Parameter estimation based on stacked regression and evolutionary algorithms

A new parameter-estimation algorithm, which minimises the cross-validated prediction error for linear-in-the-parameter models, is proposed, based on stacked regression and an evolutionary algorithm. It is initially shown that cross-validation is very important for prediction in linear-in-the-parameter models using a criterion called the mean dispersion error (MDE). Stacked regression, which can be regarded as a sophisticated type of cross-validation, is then introduced based on an evolutionary algorithm, to produce a new parameter-estimation algorithm, which preserves the parsimony of a concise model structure that is determined using the forward orthogonal least-squares (OLS) algorithm. The PRESS prediction errors are used for cross-validation, and the sunspot and Canadian lynx time series are used to demonstrate the new algorithms.

On combination of SMOTE and particle swarm optimization based radial basis function classifier for imbalanced problems

The combination of the synthetic minority oversampling technique (SMOTE) and the radial basis function (RBF) classifier is proposed to deal with classification for imbalanced two-class data. In order to enhance the significance of the small and specific region belonging to the positive class in the decision region, the SMOTE is applied to generate synthetic instances for the positive class to balance the training data set. Based on the over-sampled training data, the RBF classifier is constructed by applying the orthogonal forward selection procedure, in which the classifier structure and the parameters of RBF kernels are determined using a particle swarm optimization algorithm based on the criterion of minimizing the leave-one-out misclassification rate. The experimental results on both simulated and real imbalanced data sets are presented to demonstrate the effectiveness of our proposed algorithm.

Comparisons of host mitochondrial, nuclear and endosymbiont bacterial genes reveal cryptic fig wasp species and the effects of Wolbachia on host mtDNA evolution and diversity

Background Figs and fig-pollinating wasp species usually display a highly specific one-to-one association. However, more and more studies have revealed that the "one-to-one" rule has been broken. Co-pollinators have been reported, but we do not yet know how they evolve. They may evolve from insect speciation induced or facilitated by Wolbachia which can manipulate host reproduction and induce reproductive isolation. In addition, Wolbachia can affect host mitochondrial DNA evolution, because of the linkage between Wolbachia and associated mitochondrial haplotypes, and thus confound host phylogeny based on mtDNA. Previous research has shown that fig wasps have the highest incidence of Wolbachia infection in all insect taxa, and Wolbachia may have great influence on fig wasp biology. Therefore, we look forward to understanding the influence of Wolbachia on mitochondrial DNA evolution and speciation in fig wasps. Results We surveyed 76 pollinator wasp specimens from nine Ficus microcarpa trees each growing at a different location in Hainan and Fujian Provinces, China. We found that all wasps were morphologically identified as Eupristina verticillata, but diverged into three clades with 4.22-5.28% mtDNA divergence and 2.29-20.72% nuclear gene divergence. We also found very strong concordance between E. verticillata clades and Wolbachia infection status, and the predicted effects of Wolbachia on both mtDNA diversity and evolution by decreasing mitochondrial haplotypes. Conclusions Our study reveals that the pollinating wasp E. verticillata on F. microcarpa has diverged into three cryptic species, and Wolbachia may have a role in this divergence. The results also indicate that Wolbachia strains infecting E. verticillata have likely resulted in selective sweeps on host mitochondrial DNA.

A combined SMOTE and PSO based RBF classifier for two-class imbalanced problems

This contribution proposes a powerful technique for two-class imbalanced classification problems by combining the synthetic minority over-sampling technique (SMOTE) and the particle swarm optimisation (PSO) aided radial basis function (RBF) classifier. In order to enhance the significance of the small and specific region belonging to the positive class in the decision region, the SMOTE is applied to generate synthetic instances for the positive class to balance the training data set. Based on the over-sampled training data, the RBF classifier is constructed by applying the orthogonal forward selection procedure, in which the classifier's structure and the parameters of RBF kernels are determined using a PSO algorithm based on the criterion of minimising the leave-one-out misclassification rate. The experimental results obtained on a simulated imbalanced data set and three real imbalanced data sets are presented to demonstrate the effectiveness of our proposed algorithm.

Forecasting the future: an overview of the strategic planning of the Bournemouth and Poole conurbation

An important part of strategic planning’s purpose should be to attempt to forecast the future, not simply to belatedly respond to events, or accept the future as inevitable. This paper puts forward a conceptual approach for seeking to achieve these aims and uses the Bournemouth and Poole area in Dorset as a vehicle for applying the basic methodology. The area has been chosen because of the significant issues that it currently faces in planning terms; and its future development possibilities. In order that alternative future choices for the area – different ‘developmental trajectories’ – can be evaluated, they must be carefully and logically constructed. Four Futures for Bournemouth/Poole have been put forward; they are titled and colour-coded: Future One is Maximising Growth – Golden Prospect which seeks to achieve the highest level of economic prosperity of the area; Future Two is Incremental Growth – Solid Silver which attempts to facilitate a steady, continuing, controlled pattern of the development for the area; Future Three is Steady State – Cobalt Blue which suggests that people in the area could be more concerned with preserving their quality of life in terms of their leisure and recreation rather than increasing wealth; Future Four is Environment First – Jade Green which makes the area’s environmental protection its top priority even at the possible expense of economic prosperity. The scenarios proposed here are not sacrosanct. Nor are they simply confined to the Bournemouth and Poole area. In theory, suitably modified, they could use in a variety of different contexts. Consideration of the scenarios – wherever located - might then generate other, additional scenarios. These are called hybrids, alloys and amalgams. Likewise it might identify some of them as inappropriate or impossible. Most likely, careful consideration of the scenarios will suggest hybrid scenarios, in which features from different scenarios are combined to produce alternative or additional futures for consideration. The real issue then becomes how best to fashion such a future for the particular area under consideration

Cognitive and bodily selves: how do they interact following brain lesion?

Dualism has long distinguished between the mental and the body experiences. Probing the structure and organisation of the self traditionally calls for a distinction between these two sides of the self coin. It is far beyond the scope of this chapter to address these philosophical issues, and our starting point will be the simple distinction between reflective processes involved in the elaboration of body image, self awareness and self-recognition (i.e. ‘the self’) and the sensori-motor dialogues involved in action control, reactions and automatisms (i.e. ‘the body’ schema). This oversimplification does not take into account the complex interactions taking place between these two levels of description, but our initial aim will be to distinguish between them, before addressing the question of their interactions. Cognitive and sensori-motor processes have frequently been distinguished (review: Rossetti and Revonsuo 2000), and it may be proposed that a similar dissociation can be put forward, a priori, between a central representation of self and a bodily representation corresponding to body schema (Figure 1).

Solar cycle 24: implications for energetic particles and long-term space climate change

The recent solar minimum was the longest and deepest of the space age, with the lowest average sunspot numbers for nearly a century. The Sun appears to be exiting a grand solar maximum (GSM) of activity which has persisted throughout the space age, and is headed into a significantly quieter period. Indeed, initial observations of solar cycle 24 (SC24) continue to show a relatively low heliospheric magnetic field strength and sunspot number (R), despite the average latitude of sunspots and the inclination of the heliospheric current sheet showing the rise to solar maximum is well underway. We extrapolate the available SC24 observations forward in time by assuming R will continue to follow a similar form to previous cycles, despite the end of the GSM, and predict a very weak cycle 24, with R peaking at ∼65–75 around the middle/end of 2012. Similarly, we estimate the heliospheric magnetic field strength will peak around 6nT. We estimate that average galactic cosmic ray fluxes above 1GV rigidity will be ∼10% higher in SC24 than SC23 and that the probability of a large SEP event during this cycle is 0.8, compared to 0.5 for SC23. Comparison of the SC24 R estimates with previous ends of GSMs inferred from 9300 years of cosmogenic isotope data places the current evolution of the Sun and heliosphere in the lowest 5% of cases, suggesting Maunder Minimum conditions are likely within the next 40 years.

Fifty years of international business theory and beyond

As the field of international business has matured, there have been shifts in the core unit of analysis. First, there was analysis at country level, using national statistics on trade and foreign direct investment (FDI). Next, the focus shifted to the multinational enterprise (MNE) and the parent’s firm specific advantages (FSAs). Eventually the MNE was analysed as a network and the subsidiary became a unit of analysis. We untangle the last fifty years of international business theory using a classification by these three units of analysis. This is the country-specific advantage (CSA) and firm-specific advantage (FSA) matrix. Will this integrative framework continue to be useful in the future? We demonstrate that this is likely as the CSA/FSA matrix permits integration of potentially useful alternative units of analysis, including the broad region of the triad. Looking forward, we develop a new framework, visualized in two matrices, to show how distance really matters and how FSAs function in international business. Key to this are the concepts of compounded distance and resource recombination barriers facing MNEs when operating across national borders.

Moving feedback forward: theory to practice

There is substantial research interest in tutor feedback and students’ perception and use of such feedback. This paper considers some of the major issues raised in relation to tutor feedback and student learning. We explore some of the current feedback drivers, most notably the need for feedback to move away from simply a monologue from a tutor to a student to a valuable tutor–student dialogue. In relation to moving feedback forward the notions of self regulation, dialogue and social learning are explored and then considered in relation to how such theory can translate into practice. The paper proposes a framework (GOALS) as a tool through which tutors can move theory into practice with the aim of improving student learning from feedback.