Maximum Likelihood Estimation of Mixture Densities for Binned and Truncated Multivariate Data

Binning and truncation of data are common in data analysis and machine learning. This paper addresses the problem of fitting mixture densities to multivariate binned and truncated data. The EM approach proposed by McLachlan and Jones (Biometrics, 44: 2, 571-578, 1988) for the univariate case is generalized to multivariate measurements. The multivariate solution requires the evaluation of multidimensional integrals over each bin at each iteration of the EM procedure. Naive implementation of the procedure can lead to computationally inefficient results. To reduce the computational cost a number of straightforward numerical techniques are proposed. Results on simulated data indicate that the proposed methods can achieve significant computational gains with no loss in the accuracy of the final parameter estimates. Furthermore, experimental results suggest that with a sufficient number of bins and data points it is possible to estimate the true underlying density almost as well as if the data were not binned. The paper concludes with a brief description of an application of this approach to diagnosis of iron deficiency anemia, in the context of binned and truncated bivariate measurements of volume and hemoglobin concentration from an individual's red blood cells.

Testing insect movement theory: empirical analysis of pest data routinely collected from agricultural crops

Observations of an insect's movement lead to theory on the insect's flight behaviour and the role of movement in the species' population dynamics. This theory leads to predictions of the way the population changes in time under different conditions. If a hypothesis on movement predicts a specific change in the population, then the hypothesis can be tested against observations of population change. Routine pest monitoring of agricultural crops provides a convenient source of data for studying movement into a region and among fields within a region. Examples of the use of statistical and computational methods for testing hypotheses with such data are presented. The types of questions that can be addressed with these methods and the limitations of pest monitoring data when used for this purpose are discussed. (C) 2002 Elsevier Science B.V. All rights reserved.

Empirical evidence for ultrametric structure in multi-layer perceptron error surfaces

Combinatorial optimization problems share an interesting property with spin glass systems in that their state spaces can exhibit ultrametric structure. We use sampling methods to analyse the error surfaces of feedforward multi-layer perceptron neural networks learning encoder problems. The third order statistics of these points of attraction are examined and found to be arranged in a highly ultrametric way. This is a unique result for a finite, continuous parameter space. The implications of this result are discussed.

Factors affecting cane yield and commercial cane sugar in the Tully district

The Tully Sugar Mill has collected information about sugarcane supplied for crushing from every block in the mill district from 1970 to 1999. Data from 1988 to 1999 were analysed to understand the extent of the variation in cane yield per hectare and commercial cane sugar in the Tully mill area. The key factors influencing the variation in cane yield and commercial cane sugar in this commercial environment were identified and the variance components computed using a restricted maximum likelihood methodology. Cane yield was predominantly influenced by the year in which it was harvested, the month when the crop was ratooned (month of harvest in the previous year) and the farm of origin. These variables were relatively more important than variety, age of crop or crop class (plant crop, first ratoon through to fourth or older ratoons) and fallowing practice (fallow or ploughout-replant). The month-of-ratooning effect was relatively stable from year-to-year. Commercial cane sugar was influenced by the year of harvest, the month of harvest and their interaction, in that the influence of the month of harvest varied from year to year. Variety and farm differences were also significant but accounted for a much lower portion of the variation in commercial cane sugar. An empirical model was constructed from the key factors that influenced commercial cane sugar and cane yield to quantify their combined influence on sugar yield (t/ha). This may be used to assist mill personnel to predict their activities more accurately, for example to calculate the impact of a late finish to the current harvest season on the following year's crop.

Australian empirical study into genetic discrimination

This paper outlines a major empirical study that is being undertaken by an interdisciplinary team into genetic discrimination in Australia. The 3-year study will examine the nature and extent of this newly emerging phenomenon across the perspectives of consumers, third parties, and the legal system and will analyze its social and legal dimensions. Although the project is confined to Australia, it is expected that the outcomes will have significance for the wider research community as this is the most substantial study of its kind to be undertaken to date into genetic discrimination.

Managing new product development in the Chinese steel industry: An empirical investigation

This study explores several important aspects of the management of new product development (NPD) in the Chinese steel industry. Specifically it explores NPD success factors, the importance of management functions to new product success and measures of new product success from the perspective of the industry's practitioners. Based on a sample of 190 industrial practitioners from 18 Chinese steel companies, the study provides a mixed picture as China makes the transition from a centrally-controlled to market-based economy. On one hand, respondents ranked understanding users' needs as the most important factor influencing the performance of the new products. Further, formulating new product strategy and strengthening market research are perceived as the most important managerial functions in NPD. However, technical performance measures are regarded as more important and are more widely used in industry than market-based or financial measures of success.

Rock fracture mean trace length estimation and confidence interval calculation using maximum likelihood methods

There has been a resurgence of interest in the mean trace length estimator of Pahl for window sampling of traces. The estimator has been dealt with by Mauldon and Zhang and Einstein in recent publications. The estimator is a very useful one in that it is non-parametric. However, despite some discussion regarding the statistical distribution of the estimator, none of the recent works or the original work by Pahl provide a rigorous basis for the determination a confidence interval for the estimator or a confidence region for the estimator and the corresponding estimator of trace spatial intensity in the sampling window. This paper shows, by consideration of a simplified version of the problem but without loss of generality, that the estimator is in fact the maximum likelihood estimator (MLE) and that it can be considered essentially unbiased. As the MLE, it possesses the least variance of all estimators and confidence intervals or regions should therefore be available through application of classical ML theory. It is shown that valid confidence intervals can in fact be determined. The results of the work and the calculations of the confidence intervals are illustrated by example. (C) 2003 Elsevier Science Ltd. All rights reserved.

Modified logit life table system: Principles, empirical validation, and application

Despite its widespread use, the Coale-Demeny model life table system does not capture the extensive variation in age-specific mortality patterns observed in contemporary populations, particularly those of the countries of Eastern Europe and populations affected by HIV/AIDS. Although relational mortality models, such as the Brass logit system, can identify these variations, these models show systematic bias in their predictive ability as mortality levels depart from the standard. We propose a modification of the two-parameter Brass relational model. The modified model incorporates two additional age-specific correction factors (gamma(x), and theta(x)) based on mortality levels among children and adults, relative to the standard. Tests of predictive validity show deviations in age-specific mortality rates predicted by the proposed system to be 30-50 per cent lower than those predicted by the Coale-Demeny system and 15-40 per cent lower than those predicted using the original Brass system. The modified logit system is a two-parameter system, parameterized using values of l(5) and l(60).

Factors Influencing the Implementation of the Balanced Scorecard in Portugal: Empirical Evidence From Privately Owned and Publicly Owned Organizations

In this paper we aim to identify and analyze a set of variables that can potentially influence the adoption and knowledge of the Balanced Scorecard (BSC) in Portugal. Hypotheses were tested using data obtained from a questionnaire sent to 591 publicly-owned organizations (local governments, municipal corporations and hospitals) and 549 privately-owned organizations (large companies and small and medium enterprises) in Portugal. The results allow us to conclude that although the majority of respondents claimed to know the BSC, its use in Portugal is still limited and very recent, particularly in the public sector organizations. However, it should be noted that its use has increased in Portugal in recent years. The study also reveals that in spite of the noticeable differences between public and private sector, the BSC is used in the public sector after a few adjustments to the traditional model. Using as theoretical framework the contingency and institutional theories, we found that decentralization, vertical differentiation and the degree of higher education are associated with the implementation of the BSC.

The Knowledge and Use of the Balanced Scorecard in Portugal: Empirical Evidence of the 250 Largest Companies

Balanced Scorecard (BSC) is recognized, both in the academic and business world, as one of the most powerful strategic management accounting tools. Thus, we launched in October 2004 a questionnaire survey applied to the 250 largest Portuguese companies aiming at observing the knowledge, use, and companies’ characteristics which are adopting this management instrument. Despite the majority of the companies inquired recognize BSC more as a strategic management tool than a performance valuation system, the results show that there is still a reduced and recent utilization of BSC in Portugal. Similarly to other countries Portugal is still in the initial state of BSC utilization. Our work has shown that the companies that use more BSC belong mainly to the secondary sector of industry. Nevertheless, unlike other studies, we did not get empirical evidence on the influence of variables such as geographical localization, dimension and internationalization, in the use and knowledge of BSC in Portugal.

Disclosure of Intangible Assets: an Empirical Study of Financial Corporations in the Iberian Peninsula

The purpose of this paper is to analyse, firstly, to what extent intangible assets in the consolidated accounts of seven Portuguese banks and seven Spanish banks between 2006 and 2009 are disclosed and, secondly, to analyse what the most influential factors are in the above mentioned disclosure. In order to do this, before reviewing the existing literature and on the basis of other studies on this topic, a disclosure index has been created based on the requirements related to the intangible assets as stated in IAS 38. Then, two statistical analyses have been made: a univariate one for each of the explanatory variables and a multivariate one, in which all variables have been analysed. Both analyses led to the conclusion that the disclosure index of intangible assets is 0.96, where the bank dimension and the internationalization degree are the variables that are considered explanatory of the variation of the disclosure index in the regression analysis.