Loading...
Please wait, while we are loading the content...
Similar Documents
Easily Add Significance Testing to your Market Basket Analysis in SAS ® Enterprise Miner TM
| Content Provider | Semantic Scholar |
|---|---|
| Author | Faron, Michael Chakraborty, Goutam |
| Copyright Year | 2012 |
| Abstract | Market Basket Analysis is a popular data mining tool that can be used to search through data to find patterns of co-occurrence among objects. It is an algorithmic process that generates business rules and several metrics for each business rule such as support, confidence and lift that help researchers identify “interesting” patterns. Although useful, these popular metrics do not incorporate traditional significance testing. This paper describes how to easily add a well-known statistical significance test, the Pearson’s Chi Squared statistic, to the existing output generated by SAS® Enterprise Miner’s Association Node. The addition of this significance test enhances the ability of data analysts to make better decisions about which business rules are likely to be more useful. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://support.sas.com/resources/papers/proceedings12/204-2012.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
