October 7, 2010 at 6:47 am
Hi All
As I am very new to Data Mining.
I like to know
1. What is Data Mining.
2. How it is different from Business Intelligence.
3. What are way's to implement it.
4. How is market (Job) for the same.
Thanks
October 7, 2010 at 7:04 am
1. What is Data Mining?
Data mining can be related with self explanatory terms such as knowledge discovery, machine learning or predictive analysis. The term mining means to determine and discover patterns which are used to reason a fact. There are many data mining algorithms available which can be used to train the data.
2. How it is different from Business Intelligence?
Data mining is a bundled package which comes with Business Intelligence. It is not different from BI. Most important, BI is not a product it is a technology.
3. What are way's to implement it?
Please refer Question 1
4. How is market (Job) for the same?
The area is fairly untouched and now a days it is gaining momemtum. It is fairly interesting.
It is nice to hear that after several attempts the forum is gaining visibility.
I would recommend,if you are interested, Data Mining for SQL Server 2008 by Jamie MacLennan
Raunak J
October 7, 2010 at 7:51 am
Raunak
It's really good start here. And thanks for quick answering a these tiny question.
October 7, 2010 at 7:54 am
Pleasure.:-)
Raunak J
October 7, 2010 at 2:01 pm
Raunak Jhawar (10/7/2010)
Data mining is a bundled package which comes with Business Intelligence. It is not different from BI. Most important, BI is not a product it is a technology.
I was under the assumption that BI is a "concept" (in the sense that describes the general specifications of what is supposed to deliver) while Data Mining is a "technology" (in the sense that actually delivers what was promised by BI) - Is that a fair statement?
_____________________________________
Pablo (Paul) Berzukov
Author of Understanding Database Administration available at Amazon and other bookstores.
Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.October 7, 2010 at 9:43 pm
PaulB-TheOneAndOnly (10/7/2010)
Raunak Jhawar (10/7/2010)
Data mining is a bundled package which comes with Business Intelligence. It is not different from BI. Most important, BI is not a product it is a technology.
I was under the assumption that BI is a "concept" (in the sense that describes the general specifications of what is supposed to deliver) while Data Mining is a "technology" (in the sense that actually delivers what was promised by BI) - Is that a fair statement?
To err is human, to recurse is divine:-):-)
Raunak J
October 8, 2010 at 12:40 am
What I got is,
In BI, you already aware of patterns in Datasets.
In case of BI you have Question with you and you are looking for
Answer to that Question in the Datasets.
Ex : "Give me a breakdown of last month’s sales by salesman in the
poorest performing region" is question asked by Business, and
in response to this question you will generate some Report in BI
tool.
In Data Mining, you like to see hidden/unknown pattern in Datasets.
here you don't have Question with you.
Ex : Buying behavior of customers.
October 8, 2010 at 6:06 am
saurabh.deshpande (10/8/2010)
In Data Mining, you like to see hidden/unknown pattern in Datasets.
Let me ilustrate with a purchasing pattern discovered in a major drug store chain.
1-Pattern: "from 9pm to 6am 90% of people buying baby diapers also buy beer"
2-Root cause of the pattern: "husbands are the ones buying baby diapers at night so they take advantage of the trip to the drug store to by beer for themselves"
3-Business action: "Put beer close to the baby diapers"
There is no question business could have asked to find that.
Here is how it works...
1- Data mining shows the pattern a.k.a. correlation in between sales of two products in a specific timeframe.
2- Business investigates to find the rationale behind the pattern.
3- Business takes action to take advantage of the pattern - provided pattern is potentialy positive.
Hope this helps.
_____________________________________
Pablo (Paul) Berzukov
Author of Understanding Database Administration available at Amazon and other bookstores.
Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.October 8, 2010 at 6:07 am
Raunak Jhawar (10/7/2010)
PaulB-TheOneAndOnly (10/7/2010)
Raunak Jhawar (10/7/2010)
Data mining is a bundled package which comes with Business Intelligence. It is not different from BI. Most important, BI is not a product it is a technology.
I was under the assumption that BI is a "concept" (in the sense that describes the general specifications of what is supposed to deliver) while Data Mining is a "technology" (in the sense that actually delivers what was promised by BI) - Is that a fair statement?
To err is human, to recurse is divine:-):-)
Thank you for clarifying Raunak - I'll take that as a "yes" 🙂
_____________________________________
Pablo (Paul) Berzukov
Author of Understanding Database Administration available at Amazon and other bookstores.
Disclaimer: Advice is provided to the best of my knowledge but no implicit or explicit warranties are provided. Since the advisor explicitly encourages testing any and all suggestions on a test non-production environment advisor should not held liable or responsible for any actions taken based on the given advice.October 8, 2010 at 6:25 am
Best. Baby Diapers and Beer. Couldn't have asked a better example. You Rock. 😀
Raunak J
November 10, 2010 at 9:13 am
Raunak Jhawar (10/8/2010)
Best. Baby Diapers and Beer. Couldn't have asked a better example. You Rock. 😀
It is a classic example, used many times to explain data mining.
However, it is not true:
http://www.theregister.co.uk/2006/08/15/beer_diapers/
Seriously, have you ever been in a supermarket where beer and diapers are next to each other? 😀
Better examples I've found in every day life:
* supermarkets use data mining to see if a promotion on a product was succesfull and to see if the products in the intermediate area of said product sold less or more.
* supermarkets also use data mining to send you "personalized" coupons
* Amazon uses data mining to suggest you books that might interest you. (this is correlation. Amazon checks the books you've bought or looked at, and uses these to find books that are highly correlated with those. As more and more people come at the Amazon site, the correlation algorithm gets more and more precise.)
* analysis of the customer base to find a certain demographic that might find a certain brochure interesting
Need an answer? No, you need a question
My blog at https://sqlkover.com.
MCSE Business Intelligence - Microsoft Data Platform MVP
November 10, 2010 at 10:10 am
I was thinking about the "beers and diapers" example during dinner, and maybe there can be some truth in it:
if beer and diapers are highly correlated, you should not set them together (as in the example), but instead put them as far apart as you can (as in real life). If there's someone in the store who wants to buy beer and diapers (me, for example), he'd have to traverse the whole store. This way the store can present its wonderful products along the way, which is exactly what they want.
Need an answer? No, you need a question
My blog at https://sqlkover.com.
MCSE Business Intelligence - Microsoft Data Platform MVP
November 11, 2010 at 6:22 am
This is one test case scenario...
There are also chances of pre assumptions by a certain lot of customers pre-assuming that since many stores have a definite store arrangement they may tend to believe that the said product is not available. Which though is a negative test case...should be considered.
Raunak J
September 6, 2011 at 5:16 am
I really love the explanation so clear.
“When I hear somebody sigh, ‘Life is hard,’ I am always tempted to ask, ‘Compared to what?’” - Sydney Harris
February 6, 2012 at 7:57 am
I still don't get it... then again SSAS leaves me confused on days that end with Y
1. Say I have a well functioning cube.
A. How do I turn data mining loose on it? I realize it's part of SSAS, but making it do anything seems confusing.
B. How do I present the results of the Data Mining? Near as I can tell, the only way to do that is via the Excel Data
Mining plug-in. Except of course the Excel DM plugin doesn't work with 64-bit excel.
2. Do certain table designs work best for DM?
3. Is there a simple way to bring the DM values into the cube itself? i.e. if I'm using the DM to extrapolate a growth curve, how would I have SSAS extend the line past the end of the data?
Viewing 15 posts - 1 through 15 (of 18 total)
You must be logged in to reply to this topic. Login to reply