Tuesday, October 25, 2011

'theorem'

Bayes is a mathematical equation used in court cases to analyze statistical data. However, a judge ruled he can not be used. Does it lead to more miscarriages of justice?

not often that the silent world of mathematics is shaken by a murder case. But last summer, I saw a trial sent to academics in free fall, and has since swelled to a confrontation between science and fever law.

Basically it is a matter of chance. And it starts with a convicted murderer, "T", who took his case to the Court of Appeal in 2010. Among the evidence against him was marked by a shoe of a pair of Nike shoes, which seems to coincide with a couple in their home. While calls to discover the evidence was often weak, it was different. This time, a mathematical formula was rejected by the court. The expert in shoes made what the judge believes they were poor calculation of the possibility that the party, aggravated by a poor explanation of how he came to his opinion. The conviction was overturned.

But most importantly, what mathematicians are concerned, the judge ruled against the use of similar statistical analysis on the courts in the future. This is not the first time the judges were hostile to the use of formulas. But the real concern, forensic experts say is that failure could lead to miscarriages of justice.

"The impact will be very overwhelming," says Professor Norman Fenton, a mathematician at Queen Mary University of London. Over the last four years has been an expert witness in six cases, including the Levi Bellfield's trial, 2007 for the murder of Marsha McDonnell and Amelie Delagrange. He claims that the decision in the case of tests Imprint threat of harm to come to court because the experts because it can no longer use the mathematics they need it.

Specifically, it refers to a statistical tool called the Bayes theorem. Invented by a 18th century English mathematician, Thomas Bayes, the probability of an event occurring, given the likelihood of other related events. Some mathematicians refer to simply as logical as the Bayesian reasoning is something we do naturally. If a husband tells his wife not to eat the cake in the refrigerator, but the points of chocolate on his face, his estimate of his guilt increases. But when many factors are involved, a Bayesian calculation is more accurate for forensic scientists to measure the change in the guilt or innocence.

In the case of the murder of footprint, for example, to verify the ability to print at the crime scene came from the same pair of Nike shoes such as those found in suspect's house, given the common of this type of shoe size, how the plant had been worn or damaged. Between 1996 and 2006, for example, distributed 786,000 pairs of Nike sneakers. This could suggest that the party does not mean much. But if you consider that there are 1200 different models of Nike sneakers exclusive and sold about 42 million pairs of shoes each year, a pair becomes more important.

The data needed to perform these calculations, however, is not always available. And that's where the expert in this case was attacked. The judge complained that he could not say exactly how much a particular type of Nike Trainer in the country. The figures for domestic sales of the shoes are only rough estimates.

And he decided that Bayes' theorem should not be used again unless the underlying statistics are "enterprise". The decision could affect the traces of drugs and play clothes fibers and tests of shoes, but not DNA.

"We hope that the appeals court to reconsider this decision," said Colin Aitken, professor of forensic statistics at Edinburgh University, and President of the Working Group of the Society Royal statistical statistics and the law. It is customary, he explains, by forensic experts to use Bayes' theorem, even when the data are limited, so that the assumptions and make reasonable estimates of what figures could be. Being unable to do that, he said, could run the risk of miscarriages of justice.



Find best price for : --Lagnado----Winton----John----Fenton----David----College----Royal----Colin----University----Queen----Norman----Nike----statistical--

0 comments:

Blog Archive