Talk:Predictive analytics/Archives/2013

Survival or duration analysis
"Censoring and non-normality, which are characteristic of survival data, generate difficulty when trying to analyze the data using conventional statistical models such as multiple linear regression. The normal distribution, being a symmetric distribution, takes positive as well as negative values, but duration by its very nature cannot be negative and therefore normality cannot be assumed when dealing with duration/survival data. Hence the normality assumption of regression models is violated."

Though the normal distribution may not always be appropriate for analyzing duration or survival data, I think the reasoning above is flawed. It states that duration cannot be negative and that therefore normality cannot be assumed. However, human height, for example, cannot be negative, but no one would argue that the distribution of heights in a population cannot be normal because height cannot be negative. The author also states later in the section that duration models can be parametric, which would include the normal distribution. I'm not a statistician so I cannot offer an alternative to what has been written, but it does need to be changed. 186.176.192.3 (talk) 18:28, 2 April 2013 (UTC)

Predictive Analysis for Web Fraud
I added a link to a page on Experian's company website. Experian is a leading global information services company, providing data and analytical tools to clients around the world. The linked page shows recent advancements in technology have also introduced predictive behavior analysis for web fraud detection. (talk) 22:01, 27 October 2013 (UTC)
 * I removed it as an example and primary source that comes across as advertising. We need an independent and secondary/tertiary source instead. --Ronz (talk) 15:53, 28 October 2013 (UTC)