When model training, algorithm read whole training data stored in database.
If whole training data is not fit in memory, how can algorithm handle this situation?
Is it disk based algorithm?
Or does it use sampling or something special techniques?
NOTICE: developerWorks Community will be offline May 29-30, 2015 while we upgrade to the latest version of IBM Connections. For more information, read our upgrade FAQ.
This topic has been locked.
2 replies Latest Post - 2011-04-15T12:50:58Z by Sprince
Pinned topic About data mining algorithm supported by intelligent miner...
Answered question This question has been answered.
Unanswered question This question has not been answered yet.
Updated on 2011-04-15T12:50:58Z at 2011-04-15T12:50:58Z by Sprince
M._Haide 060000132W6 PostsACCEPTED ANSWER
Re: About data mining algorithm supported by intelligent miner...2011-04-15T12:00:40Z in response to SprinceHi Sprince,
there are different methods to handle such situation. Depending on the algorithm you select IM does use disk space in case the available memory is not enough.
If you like to use sampling you need to specify it otherwise all data is processed.