Estimation of the bayes' factor in a forensic science problem

Chan, Karen; Aitken, Colin

doi:10.1080/00949658908811201

Cited by 16 publications

(9 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Earlier examples of the use of the kernel distribution in forensic science are Aitken (1986, 1995), Evett et al . (1987), Chan and Aitken (1989), Berry (1991) and Berry et al . (1992).…”

Section: 5 Likelihood Ratio Using a Kernel Distribution For The DImentioning

confidence: 94%

“…The assumption of normality can be removed by considering a kernel density estimate for the between-group distribution. Earlier examples of the use of the kernel distribution in forensic science are Aitken (1986Aitken ( , 1995, Evett et al (1987), Chan and Aitken (1989), Berry (1991) and Berry et al (1992). Given a data set, which in this case will be taken to be the group means .x 1 , .…”

Section: Likelihood Ratio Using a Kernel Distribution For The Distribmentioning

confidence: 99%

See 1 more Smart Citation

Evaluation of Trace Evidence in the Form of Multivariate Data

Aitken

Lucy

2004

Journal of the Royal Statistical Society Series C: Applied Statistics

183

234

View full text Add to dashboard Cite

The evaluation of measurements on characteristics of trace evidence found at a crime scene and on a suspect is an important part of forensic science. Five methods of assessment for the value of the evidence for multivariate data are described. Two are based on significance tests and three on the evaluation of likelihood ratios. The likelihood ratio which compares the probability of the measurements on the evidence assuming a common source for the crime scene and suspect evidence with the probability of the measurements on the evidence assuming different sources for the crime scene and suspect evidence is a well-documented measure of the value of the evidence. One of the likelihood ratio approaches transforms the data to a univariate projection based on the first principal component. The other two versions of the likelihood ratio for multivariate data account for correlation among the variables and for two levels of variation: that between sources and that within sources. One version assumes that between-source variability is modelled by a multivariate normal distribution; the other version models the variability with a multivariate kernel density estimate. Results are compared from the analysis of measurements on the elemental composition of glass. Copyright 2004 Royal Statistical Society.

show abstract

“…Earlier examples of the use of the kernel distribution in forensic science are Aitken (1986, 1995), Evett et al . (1987), Chan and Aitken (1989), Berry (1991) and Berry et al . (1992).…”

Section: 5 Likelihood Ratio Using a Kernel Distribution For The DImentioning

confidence: 94%

Section: Likelihood Ratio Using a Kernel Distribution For The Distribmentioning

confidence: 99%

Evaluation of Trace Evidence in the Form of Multivariate Data

Aitken

Lucy

2004

Journal of the Royal Statistical Society Series C: Applied Statistics

183

234

View full text Add to dashboard Cite

show abstract

“…Pettit (1990) showed that the observation with minimum CPO lies at one of the vertices of the convex hull of the observations and gave results in which the data are assumed to follow the multivariate normal distribution, the mean of the normal distribution is assumed to have a normal distribution and, if the variance is not assumed known, the covariance matrix has a Wishart distribution. Chan and Aitken (1989) and Berry et 01. (1992) considered problems in which the assumption of a normal prior distribution was not thought realistic and used a nonparametric estimate of the prior distribution.…”

Section: Conditional Predictive Ordinatementioning

confidence: 92%

Contaminant Detection in the Visual Inspection of Seed Samples

1995

View full text Add to dashboard Cite

The quality of a seed lot is determined, in part, by the amount of contaminants in a sample. One way of automating the process of identifying contaminants is to take shape and size measurements of each item in the sample. The process of detecting contaminants from such data may be viewed as a multivariate statistical outlier problem in which contaminants are considered as outliers in a sample of normal seeds. Pettit has developed Bayesian diagnostics for multivariate normal distributions with multivariate normal prior distributions; an extension is described here in which the prior distribution for the mean is other than normal and is represented by a kernel density estimate. The performances of the normal distribution and extended methods are compared in an application to the seed testing problem.

show abstract

“…Earlier examples of the use of the KDE in forensic science are available in Aitken (1986), Aitken and Taroni (2004), Berry (1991), Berry et al (1992), Chan and Aitken (1989), and Evett et al (1987). the distribution of the data within a database) may not always be correct.…”

Section: Between-object Distribution Modelled By Kernel Density Estimmentioning

confidence: 99%