Influence functions of two families of robust estimators under proportional scatter matrices

In this paper, under a proportional model, two families of robust estimates for the proportionality constants, the common principal axes and their size are discussed. The first approach is obtained by plugging robust scatter matrices on the maximum likelihood equations for normal data. A projection- pursuit and a modified projection-pursuit approach, adapted to the proportional setting, are also considered. For all families of estimates, partial influence functions are obtained and asymptotic variances are derived from them. The performance of the estimates is compared through a Monte Carlo study. © 2006 Springer-Verlag.

Demographic demand systems with application to equivalence scales estimation and inequality analysis : the Australian evidence

This paper proposes and applies an alternative demographic procedure for extending a demand system to allow for the effect of household size and composition changes, along with price changes, on expenditure allocation. The demographic procedure is applied to two recent demand functional forms to obtain their estimable demographic extensions. The estimation on pooled time series of Australian Household Expenditure Surveys yields sensible and robust estimates of the equivalence scale, and of its variation with relative prices. Further evidence on the usefulness of this procedure is provided by using it to evaluate the nature and magnitude of the inequality bias of relative price changes in Australia over a period from the late 1980s to the early part of the new millennium.

A novel bio-kinematic encoder for human exercise representation and decomposition - Part 1 : Indexing and modelling

Current bio-kinematic encoders use velocity, acceleration and angular information to encode human exercises. However, in exercise physiology there is a need to distinguish between the shape of the trajectory and its execution dynamics. In this paper we propose such a two-component model and explore how best to compute these components of an action. In particular, we show how a new spatial indexing scheme, derived directly from the underlying differential geometry of curves, provides robust estimates of the shape and dynamics compared to standard temporal indexing schemes.

Food demand elasticities for Australia

There is renewed interest in robust estimates of food demand elasticities at a disaggregated level not only to analyse the impact of changing food preferences on the agricultural sector, but also to establish the likely impact of pricing incentives on households. Using data drawn from two national Household Expenditure Surveys covering the periods 1998/1999 and 2003/2004, and adopting an Almost Ideal Demand System approach that addresses the zero observations problem, this paper estimates a food demand system for 15 food categories for Australia. The categories cover the standard food items that Australian households demand routinely. Own-price, cross-price and expenditure elasticity estimates of the Marshallian and Hicksian types have been derived for all categories. The parameter estimates obtained in this study represent the first integrated set of food demand elasticities based on a highly disaggregated food demand system for Australia, and all accord with economic intuition.

Using repeatability to study physiological and behavioural traits: ignore time-related change at your peril

© 2015 The Association for the Study of Animal Behaviour. Broad sense repeatability, which refers to the extent to which individual differences in trait scores are maintained over time, is of increasing interest to researchers studying behavioural or physiological traits. Broad sense repeatability is most often inferred from the statistic R (the intraclass correlation, or narrow sense repeatability). However, R ignores change over time, despite the inherent longitudinal nature of the data (repeated measures over time). Here, we begin by showing that most studies ignore time-related change when estimating broad sense repeatability, and estimate R with low statistical power. Given this problem, we (1) outline how and why ignoring time-related change in scores (that occurs for whatever reason) can seriously affect estimates of the broad sense repeatability of behavioural or physiological traits, (2) discuss conditions in which various indices of R can or cannot provide reliable estimates of broad sense repeatability, and (3) provide suggestions for experimental designs for future studies. Finally, given that we already have abundant evidence that many labile traits are 'repeatable' in that broad sense (i.e. R>. 0), we suggest a shift in focus towards obtaining robust estimates of the repeatability of behavioural and physiological traits. Given how labile these traits are, this will require greater experimental (and/or statistical) control and larger sample sizes in order to detect and quantify change over time (if present).

The effects of clustered data on standard error estimates in covariance structure analysis : a field data application

Purpose – The purpose of this article is to present an empirical analysis of complex sample data with regard to the biasing effect of non-independence of observations on standard error parameter estimates. Using field data structured in the form of repeated measurements it is to be shown, in a two-factor confirmatory factor analysis model, how the bias in SE can be derived when the non-independence is ignored.

Design/methodology/approach – Three estimation procedures are compared: normal asymptotic theory (maximum likelihood); non-parametric standard error estimation (naïve bootstrap); and sandwich (robust covariance matrix) estimation (pseudo-maximum likelihood).

Findings – The study reveals that, when using either normal asymptotic theory or non-parametric standard error estimation, the SE bias produced by the non-independence of observations can be noteworthy.

Research limitations/implications – Considering the methodological constraints in employing field data, the three analyses examined must be interpreted independently and as a result taxonomic generalisations are limited. However, the study still provides “case study” evidence suggesting the existence of the relationship between non-independence of observations and standard error bias estimates.

Originality/value – Given the increasing popularity of structural equation models in the social sciences and in particular in the marketing discipline, the paper provides a theoretical and practical insight into how to treat repeated measures and clustered data in general, adding to previous methodological research. Some conclusions and suggestions for researchers who make use of partial least squares modelling are also drawn.

Robust traffic classification with mislabelled training samples

Traffic classification plays the significant role in the network security and management. However, accurate classification is challenging if the training data is contaminated with unclean traffic. Recent researches often assume clean training data, and hence performance reduced on real-time network traffic. To meet this challenge, in this paper, we propose a robust method, Unclean Traffic Classification (UTC), which incorporates noise elimination and suspected noise reweighting. Firstly, UTC eliminates strong noisy training data identified by a consensus filtering with multiple classifiers. Furthermore, UTC estimates the relevance of remaining training data and learns a robust traffic classifier. Through a number of experiments on a real-world traffic dataset, we show that the new method outperforms existing state-of-the-art traffic classification methods, under the extremely difficult circumstance with unclean training data.

Integrating acoustic telemetry into mark-recapture models to improve the precision of apparent survival and abundance estimates

Capture-mark-recapture models are useful tools for estimating demographic parameters but often result in low precision when recapture rates are low. Low recapture rates are typical in many study systems including fishing-based studies. Incorporating auxiliary data into the models can improve precision and in some cases enable parameter estimation. Here, we present a novel application of acoustic telemetry for the estimation of apparent survival and abundance within capture-mark-recapture analysis using open population models. Our case study is based on simultaneously collecting longline fishing and acoustic telemetry data for a large mobile apex predator, the broadnose sevengill shark (Notorhynchus cepedianus), at a coastal site in Tasmania, Australia. Cormack-Jolly-Seber models showed that longline data alone had very low recapture rates while acoustic telemetry data for the same time period resulted in at least tenfold higher recapture rates. The apparent survival estimates were similar for the two datasets but the acoustic telemetry data showed much greater precision and enabled apparent survival parameter estimation for one dataset, which was inestimable using fishing data alone. Combined acoustic telemetry and longline data were incorporated into Jolly-Seber models using a Monte Carlo simulation approach. Abundance estimates were comparable to those with longline data only; however, the inclusion of acoustic telemetry data increased precision in the estimates. We conclude that acoustic telemetry is a useful tool for incorporating in capture-mark-recapture studies in the marine environment. Future studies should consider the application of acoustic telemetry within this framework when setting up the study design and sampling program.