12 when to use data mining
TRANSCRIPT
![Page 1: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/1.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 1/19
When to use Data Mining
![Page 2: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/2.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 2/19
Introduction
� An important question that should be answered before youcommence any data mining project is whether data miningtechniques are, in fact necessary.
� In determining this it is important to understand what levelof sophistication of data mining is required. For instance,do you just need a few standardized printed reports or doyou need interactive ROI analysis or OLAP analysis to seewhat your data looks like?
� Do you need or true data mining techniques that buildpredictive models to search through your database for useful patterns?
![Page 3: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/3.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 3/19
The Data Mining Process
![Page 4: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/4.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 4/19
![Page 5: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/5.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 5/19
What all Data Mining techniques
have in common� Each Data Mining algorithm has the following in
common:
± Model Structure. The structure that defines the model(Is it a tree, a neural network, or a neighbor?)
± Search. How does the algorithm amend and modify the
model over time as more data is made available
± Validation. When does the algorithm terminate becauseit has created a valid model?
![Page 6: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/6.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 6/19
What all Data Mining techniqueshave in common (cont¶d)
![Page 7: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/7.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 7/19
![Page 8: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/8.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 8/19
Data Mining in the Business Process� When Data Mining is used for non-exploratory
reasons or whenever supervised learning
techniques are used, this customer reactionprovide a fairly well-defined target column withinthe database, which relates to the business process.The target must have the following attributes inorder to be successful with data mining:
± The target has value
± The target is actionable
± The effect of action can be captured
![Page 9: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/9.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 9/19
Data Mining in the Business Process (cont¶d)
![Page 10: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/10.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 10/19
![Page 11: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/11.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 11/19
Avoiding some big mistakes in Data Mining
� The technology-centered view of the data
mining process emphasizes getting the
model right, with the assumption that the
predictive product has been well-defined
and that the data that has been captured to
date is well understood.� This is not always the case.
![Page 12: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/12.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 12/19
![Page 13: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/13.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 13/19
![Page 14: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/14.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 14/19
T
hree measures for Data MiningT
ools
� Accuracy. The data mining tool must produce a model thatis as accurate as possible.
� Explanation. The data mining tool needs to be able toµexplain¶ how the model works to the end user in a clear way
� Integration. The data mining tool must integrate with thecurrent business process, and data and information flow in
the company.� When these three requirements are well met, the datamining tools will produce highly profitable models that arelikely to remain stable over long periods of time.
![Page 15: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/15.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 15/19
Embedded Data Mining for business
![Page 16: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/16.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 16/19
![Page 17: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/17.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 17/19
How to measure Accuracy,
Explanation, and Integration� Measuring Accuracy:
± Accuracy
± Error rate
± Error rate at rejection
± Mean squared error
± Lift
± Profit/ROI
![Page 18: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/18.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 18/19
How to measure Accuracy,
Explanation, and Integration� Measuring Explanation:
± Automated rule generation
± OLAP integration
± Model validation
� Measuring Integrity
± Proprietary data extracts
± Metadata
± Predictor preprocessing
± Predictor/prediction types
± Dirty data
± Missing values
± Scalability
![Page 19: 12 When to Use Data Mining](https://reader031.vdocuments.site/reader031/viewer/2022021214/577d2d2a1a28ab4e1ead01f0/html5/thumbnails/19.jpg)
8/7/2019 12 When to Use Data Mining
http://slidepdf.com/reader/full/12-when-to-use-data-mining 19/19
What the Future holds for
Embedded Data Mining� Once the data mining process becomes easy enough to use
and is seamlessly integrated into business process and the
general data and information flow around the enterprise,
there will be new applications and synergies that will make
data mining an even more critical requirement for any fully
functioning data warehouse
± Use data mining to improve the multidimensional database
± Use data mining to improve the data warehouse structure± Multidimensional databases and summary data will enhance data
mining performance. The more data, the better any data mining
technique is