nfung
October 2nd, 2002, 04:33 PM
What's in OLAP that's you cant do with simple EXCEL? I'm kind of newbie in data mining, but i have the feeling that there's two main area to it:
data mining = data collection + data preparation + analysis
(a) Data collection = data entry, replication/logshipping...etc (which is prolly the most important part since what u pick to monitor as input variables is prolly the most significant factor in determing how well your model describes the system behavoir..)
(b) Data Prep = DTS, data cleansing and preliminary manipulation
(c) analysis = report, graphing, data modeling (regression models, statistics...etc), generating the OLAP cube...
Is this correct so far? But i dont see why u cant just build a view and export the selected data to a flat file, then export it to Excel for graphing and stuff...., or use SPSS to do regression/data modeling tasks?
I am kind of newbie and have no idea of an OLAP cube is yet. And i dont see why data mining is so complex....? Reading up on Sam's OLAP unleashed, read a few chapters, done the DTS packages and i know SQL server also... but still cant see what can be done with OLAP cannot be done with Excel/SPSS/Mathematica regarding the graphing and modeling - part (except the maximum number of records of course)
seems to me that OLAP is only good at data collection and preparation....? or may be i dont understand this fully?
data mining = data collection + data preparation + analysis
(a) Data collection = data entry, replication/logshipping...etc (which is prolly the most important part since what u pick to monitor as input variables is prolly the most significant factor in determing how well your model describes the system behavoir..)
(b) Data Prep = DTS, data cleansing and preliminary manipulation
(c) analysis = report, graphing, data modeling (regression models, statistics...etc), generating the OLAP cube...
Is this correct so far? But i dont see why u cant just build a view and export the selected data to a flat file, then export it to Excel for graphing and stuff...., or use SPSS to do regression/data modeling tasks?
I am kind of newbie and have no idea of an OLAP cube is yet. And i dont see why data mining is so complex....? Reading up on Sam's OLAP unleashed, read a few chapters, done the DTS packages and i know SQL server also... but still cant see what can be done with OLAP cannot be done with Excel/SPSS/Mathematica regarding the graphing and modeling - part (except the maximum number of records of course)
seems to me that OLAP is only good at data collection and preparation....? or may be i dont understand this fully?