Objective measures for association pattern analysis
Michael S. Steinbach, Pang‐Ning Tan, Hui Yun Xiong, Vipin Kumar · Contemporary mathematics - American Mathematical Society · 2007
Data mining is an area of data analysis that has arisen in response to new data analysis challenges, such as those posed by massive data sets or non-traditional types of data. Association analysis, which seeks to find pat- terns that describe the relationships of attributes (variables) in a binary data set, is an area of data mining that has created a unique set of data analysis tools and concepts that have been widely employed in business and science. The ob- jective measures used to evaluate the interestingness of association patterns are a key aspect of association analysis. Indeed, different objective measures define different association patterns with different properties and applications. This paper first provides a general discussion of objective measures for assessing the interestingness of association patterns. It then illustrators the importance of choosing the appropriate measure for a particular application and type of data through an example that focuses on one of these measures, h-confidence, which is appropriate for binary data sets with skewed distributions. The usefulness of h-confidence and the association pattern that it defines—a hyperclique— is illustrated by an application that involves finding functional modules from protein complex data.