Is predictive analytics for marketers really that accurate?
暂无分享,去创建一个
Developing mathematical models does convey an impression or perception to the average business person that these tools are highly accurate due to the complexity in creating these tools. This article attempts to dispel this notion by conveying to the reader the accuracy limitations in developing predictive models. At the same time, the author discusses some of the rationale as to why these limitations exist. Yet, even with model accuracy being a limiting factor, these tools still yield tremendous business benefits that accrue right to the bottom line.
[1] R. Real,et al. AUC: a misleading measure of the performance of predictive distribution models , 2008 .
[2] K. Zou,et al. Receiver-Operating Characteristic Analysis for Evaluating Diagnostic Tests and Predictive Models , 2007, Circulation.
[3] David R. B. Stockwell,et al. Effects of sample size on accuracy of species distribution models , 2002 .