Precision-recall operating characteristic (P-ROC) curves in imprecise environments

Traditionally, machine learning algorithms have been evaluated in applications where assumptions can be reliably made about class priors and/or misclassification costs. In this paper, we consider the case of imprecise environments, where little may be known about these factors and they may well vary significantly when the system is applied. Specifically, the use of precision-recall analysis is investigated and compared to the more well known performance measures such as error-rate and the receiver operating characteristic (ROC). We argue that while ROC analysis is invariant to variations in class priors, this invariance in fact hides an important factor of the evaluation in imprecise environments. Therefore, we develop a generalised precision-recall analysis methodology in which variation due to prior class probabilities is incorporated into a multi-way analysis of variance (ANOVA). The increased sensitivity and reliability of this approach is demonstrated in a remote sensing application.

Building business cases for maintenance improvement projects

Many maintenance managers find it difficult to justify investments in maintenance improvement initiatives. In part, this is due to a tendency by mine managers to regard maintenance purely as a cost centre, and not as a process able to influence productive capacity and profit. It is also hindered by a lack of alignment between commonly used maintenance performance measures and key business drivers, and the lack of formal business training amongst maintenance professionals. With this in mind, a model to assist maintenance managers in evaluating the benefits of maintenance improvement projects was recently formulated. The model considers four cost saving dimensions. These are: 1. reduction in the cost of unplanned repairs and maintenance, 2. increased or accelerated production and/or sales, 3. spares inventory reduction, and 4. reduction in over-investment in physical assets and operating costs. This paper discusses the application of this model and a number of numerical examples are given to justify investments in maintenance improvement projects having varying objectives.

General gaussian priors for improved generalisation

We explore the dependence of performance measures, such as the generalization error and generalization consistency, on the structure and the parameterization of the prior on `rules', instanced here by the noisy linear perceptron. Using a statistical mechanics framework, we show how one may assign values to the parameters of a model for a `rule' on the basis of data instancing the rule. Information about the data, such as input distribution, noise distribution and other `rule' characteristics may be embedded in the form of general gaussian priors for improving net performance. We examine explicitly two types of general gaussian priors which are useful in some simple cases. We calculate the optimal values for the parameters of these priors and show their effect in modifying the most probable, MAP, values for the rules.

Dynamics of learning with restricted training sets

The dynamics of supervised learning in layered neural networks were studied in the regime where the size of the training set is proportional to the number of inputs. The evolution of macroscopic observables, including the two relevant performance measures can be predicted by using the dynamical replica theory. Three approximation schemes aimed at eliminating the need to solve a functional saddle-point equation at each time step have been derived.

Binocular interaction:Contrast matching and contrast discrimination are predicted by the same model

How do signals from the 2 eyes combine and interact? Our recent work has challenged earlier schemes in which monocular contrast signals are subject to square-law transduction followed by summation across eyes and binocular gain control. Much more successful was a new 'two-stage' model in which the initial transducer was almost linear and contrast gain control occurred both pre- and post-binocular summation. Here we extend that work by: (i) exploring the two-dimensional stimulus space (defined by left- and right-eye contrasts) more thoroughly, and (ii) performing contrast discrimination and contrast matching tasks for the same stimuli. Twenty-five base-stimuli made from 1 c/deg patches of horizontal grating, were defined by the factorial combination of 5 contrasts for the left eye (0.3-32%) with five contrasts for the right eye (0.3-32%). Other than in contrast, the gratings in the two eyes were identical. In a 2IFC discrimination task, the base-stimuli were masks (pedestals), where the contrast increment was presented to one eye only. In a matching task, the base-stimuli were standards to which observers matched the contrast of either a monocular or binocular test grating. In the model, discrimination depends on the local gradient of the observer's internal contrast-response function, while matching equates the magnitude (rather than gradient) of response to the test and standard. With all model parameters fixed by previous work, the two-stage model successfully predicted both the discrimination and the matching data and was much more successful than linear or quadratic binocular summation models. These results show that performance measures and perception (contrast discrimination and contrast matching) can be understood in the same theoretical framework for binocular contrast vision. © 2007 VSP.

The role of market orientation in the business success of the MNC’s foreign subsidiaries

In recent discussions over the contribution of marketing to the strategy dialogue, market orientation has been singled out as being of particular importance in relation to the understanding of competitive advantage (Day et al 1992, Hunt and Lamb 2000). Research in the past has focused primarily on firms operating in domestic markets. As such, despite the recent progress, it is unclear of relevancy of market orientation as a construct in the context of multinational corporations (MNC) and their foreign subsidiaries. In this study, we set out to explore the role of market orientation in the subsidiary business performance. An investigation of a sample of 252 foreign subsidiaries in the UK revealed that except for “receptive? subsidiaries (Taggart 1998), market orientation has significant positive relationships with a number of business performance measures in all three other types of subsidiaries, suggesting that market orientation is a key driver for business performance at foreign subsidiaries.

Shining in the center:central gaze cascade effect on product choice

Consumers' tendency to choose the option in the center of an array and the process underlying this effect is explored. Findings from two eye-tracking studies suggest that brands in the horizontal center receive more visual attention. They are more likely to be chosen. Investigation of the attention process revealed an initial central fixation bias, a tendency to look first at the central option, and a central gaze cascade effect, progressively increasing attention focused on the central option right prior to decision. Only the central gaze cascade effect was related to choice. An offline study with tangible products demonstrated that the centrally located item within a product category is chosen more often, even when it is not placed in the center of the visual field. Despite widespread use, memory-based attention measures were not correlated with eye-tracking measures. They did not capture visual attention and were not related to choice. © 2012 by JOURNAL OF CONSUMER RESEARCH, Inc.

Individual differences in work behaviour of blue-collar workers : a study of rubber tappers in Malaysia

This thesis examines individual differences in work behaviour of rubber tappers. The study examined sex, age, experience and race differences and their interactions with terrain on job performance, absenteeism, and job satisfaction of 1053 rubber tappers. Rubber tappers are unskilled blue-collar workers who essentially do the same type of work and are paid the same rates of pay. There are very few studies that have compared male and female blue-collar workers doing similar jobs in organisational settings. This study is one of the few investigations that examine sex differences in job performance of blue-collar workers doing same job using production data. Studies on age differences in work behaviour encounter numerous methodological difficulties such as high turnover, internal transfers and problems associated with age differences in educational levels. The participation of rubber tappers in this study is envisaged to overcome these difficulties because attrition rates of rubber tappers are low, and internal transfers are non existent. Further, the educational levels of rubber tappers are relatively similar across different age cohorts, as most rubber tappers have little or no education. Two measures of both job performance and absenteeism were derived from payroll records. The two job performance measures were total crop production and attendance. The two absenteeism measures were avoidable and unavoidable absence rates. Overall job satisfaction was determined using a 4-item scale. Significant sex, age, experience and race differences were obtained for job performance, absenteeism and job satisfaction. Significant interactive effects were also obtained for sex, age , experience, race and terrain for job performance and absenteeism. The results are discussed in relation to the abilities and motivation of rubber tappers. The implication of these findings for employee selection and human resource management in rubber estates is discussed.

The human factors of computer input using a mouse device

This thesis investigates how people select items from a computer display using the mouse input device. The term computer mouse refers to a class of input devices which share certain features, but these may have different characteristics which influence the ways in which people use the device. Although task completion time is one of the most commonly used performance measures for input device evaluation, there is no consensus as to its definition. Furthermore most mouse studies fail to provide adequate assurances regarding its correct measurement.Therefore precise and accurate timing software were developed which permitted the recording of movement data which by means of automated analysis yielded the device movements made. Input system gain, an important task parameter, has been poorly defined and misconceptualized in most previous studies. The issue of gain has been clarified and investigated within this thesis. Movement characteristics varied between users and within users, even for the same task conditions. The variables of target size, movement amplitude, and experience exerted significant effects on performance. Subjects consistently undershot the target area. This may be a consequence of the particular task demands. Although task completion times indicated that mouse performance had stabilized after 132 trials the movement traces, even of very experienced users, indicated that there was still considerable room for improvement in performance, as indicated by the proportion of poorly made movements. The mouse input device was suitable for older novice device users, but they took longer to complete the experimental trials. Given the diversity and inconsistency of device movements, even for the same task conditions, caution is urged when interpreting averaged grouped data. Performance was found to be sensitive to; task conditions, device implementations, and experience in ways which are problematic for the theoretical descriptions of device movement, and limit the generalizability of such findings within this thesis.

Corporate restructuring: empirical evidence from the UK

Corporate restructuring is perceived as a challenge to research. Prior studies do not provide conclusive evidence regarding the effects of restructuring. Since there are discernible findings, this research attempts to examine the effects of restructuring events amongst the UK listed firms. The sample firms are listed in the LSE and London AIM stock exchange. Only completed restructuring transactions are included in the study. The time horizon extends from year 1999 to 2003. A three-year floating window is assigned to examine the sample firms. The key enquiry is to scrutinise the ex post effects of restructuring on performance and value measures of firms with contrast to a matched criteria non-restructured sample. A cross sectional study employing logit estimate is undertaken to examine firm characteristics of restructuring samples. Further, additional parameters, i.e. Conditional Volatility and Asymmetry are generated under the GJR-GARCH estimate and reiterated in logit models to capture time-varying heteroscedasticity of the samples. This research incorporates most forms of restructurings, while prior studies have examined certain forms of restructuring. Particularly, these studies have made limited attempts to examine different restructuring events simultaneously. In addition to logit analysis, an event study is adopted to evaluate the announcement effect of restructuring under both the OLS and GJR-GARCH estimate supplementing our prior results. By engaging a composite empirical framework, our estimation method validates a full appreciation of restructuring effect. The study provides evidence that restructurings indicate non-trivial significant positive effect. There are some evidences that the response differs because of the types of restructuring, particularly while event study is applied. The results establish that performance measures, i.e. Operating Profit Margin, Return on Equity, Return on Assets, Growth, Size, Profit Margin and Shareholders' Ownership indicate consistent and significant increase. However, Leverage and Asset Turn Over suggest reasonable influence on restructuring across the sample period. Similarly, value measures, i.e. Abnormal Returns, Return on Equity and Cash Flow Margin suggest sizeable improvement. A notable characteristic seen coherently throughout the analysis is the decreasing proportion of Systematic Risk. Consistent with these findings, Conditional Volatility and Asymmetry exhibit similar trend. The event study analysis suggests that on an average market perceives restructuring favourably and shareholders experience significant and systematic positive gain.

Interval timing in children:effects of auditory and visual pacing stimuli and relationships with reading and attention variables

Motor timing tasks have been employed in studies of neurodevelopmental disorders such as developmental dyslexia and ADHD, where they provide an index of temporal processing ability. Investigations of these disorders have used different stimulus parameters within the motor timing tasks which are likely to affect performance measures. Here we assessed the effect of auditory and visual pacing stimuli on synchronised motor timing performance and its relationship with cognitive and behavioural predictors that are commonly used in the diagnosis of these highly prevalent developmental disorders. Twenty- one children (mean age 9.6 years) completed a finger tapping task in two stimulus conditions, together with additional psychometric measures. As anticipated, synchronisation to the beat (ISI 329 ms) was less accurate in the visually paced condition. Decomposition of timing variance indicated that this effect resulted from differences in the way that visual and auditory paced tasks are processed by central timekeeping and associated peripheral implementation systems. The ability to utilise an efficient processing strategy on the visual task correlated with both reading and sustained attention skills. Dissociations between these patterns of relationship across task modality suggest that not all timing tasks are equivalent.

Paving the way for "distinguished marketing"

Over the last six decades, marketing concepts, tools, and knowledge have gone through tremendous developments. A general trend toward formalization has affected orientation and decision making and has clarified the relationship between marketing efforts and performance measures. This evolution has received strong support from concurrent revolutions in data collection and research techniques. This article outlines the formalization of the marketing discipline and proposes steps that will pave the way for future developments in marketing, toward what I call “distinguished marketing”.

What does "good" look like? A three-country examination of marketing practice

Grounded in configuration theory, this study investigates the notion of co-alignment of business orientation, marketing assets and marketing capabilities, and their relationships to performance. Using these criteria, profiles of high performing businesses were derived and assessed against a three country sample of Brazil, China and the UK. Findings are consistent, statistically significant and invariant across the sample. They show that businesses with ideal profiles significantly outperform competitors in terms of market-based performance, customer satisfaction, and financial performance. Furthermore, profiles of top performing organizations are similar across countries with respect to their orientations, assets, and capabilities. Only customer-based assets, network capabilities, and customer and shareholder orientations were different. Implications and future research directions are subsequently addressed.

Assessing walking limitations in stroke survivors:are self-reports and proxy-reports interchangeable?

Objective: To examine the extent to which proxy-report measures adequately assess walking limitations and are interchangeable with self-report measures in stroke survivors. Design and Participants: Self-report, proxy-report, and observed performance measures of walking limitations were compared cross-sectionally on 3 occasions following the discharge from hospital of 101 stroke survivors. Correlations between measures, differences between mean scores, and agreement of self- and proxy reports were analyzed. Results and Conclusions: Self- and proxy-report measures correlated significantly with each other and with observed performance measures; differences between mean scores were not found. Agreement between individual self- and proxy-report pairs was poor, however, indicating that freely substituting proxy-report data for self-report data when self-report data are unavailable is inappropriate.

Evaluation of Pareto/D/1/k Queue by Simulation

The finding that Pareto distributions are adequate to model Internet packet interarrival times has motivated the proposal of methods to evaluate steady-state performance measures of Pareto/D/1/k queues. Some limited analytical derivation for queue models has been proposed in the literature, but their solutions are often of a great mathematical challenge. To overcome such limitations, simulation tools that can deal with general queueing system must be developed. Despite certain limitations, simulation algorithms provide a mechanism to obtain insight and good numerical approximation to parameters of queues. In this work, we give an overview of some of these methods and compare them with our simulation approach, which are suited to solve queues with Generalized-Pareto interarrival time distributions. The paper discusses the properties and use of the Pareto distribution. We propose a real time trace simulation model for estimating the steady-state probability showing the tail-raising effect, loss probability, delay of the Pareto/D/1/k queue and make a comparison with M/D/1/k. The background on Internet traffic will help to do the evaluation correctly. This model can be used to study the long- tailed queueing systems. We close the paper with some general comments and offer thoughts about future work.