What is the best practice in setting a minimum support threshold for market basket analysis Is there a formula Does it depend on ROI you predict
Setting a Minimum support threshold for association rule mining
Hot Topic
- query large database
- Problem starting and using SQL Server Mangement Studio
- The task Transfer SQL Server Objects Task cannot run on this edition of Integration Services. It requires higher level edition.
- joiing two unrelated tables?
- loading mining model error
- Which windows service will be running on my machine if I installed SDE
- MCTS Exam 70-431 Database Administration Study Documementation and Samples ?
- Save SQL Query to XML
- Login failed connecting to new principal after failover
- Server Backup & Restore
SQL Server
- VS2005/SSR2005 Report Controlling field visibility by code
- An attempt has been made to use a data extension 'DLLName' that is not registered for this report server.
- Execution of Package failed
- Indexing XML
- Programmatically creating and executing data driven subscriptions in SSRS
- SQL connectionstring parameters
- SSIS - Data conversion failed - "The value could not be converted because of a potential loss of data."
- Reporting Services 2005 and .NET 2003
- can i restore only a table from a backup file?
- Architectural Design Question ?
Setting a Minimum support threshold for association rule mining
Kolf
Nick Darnell
There isn't a formula. The minimum threshold is there for two reasons. One, to filter out data that is inconsequential, and two, to prevent the algorithm from running out of memory.
If you only want to find strong correlations, you can set it high. If you want a wide range of unknown correlations, set it low. It's best to experiment with the results and performance you get from the algorithm to see which settings work best for your particular situation.
MaartenClaes