Data Mining Techniques

Data mining concepts and techniques

1. Md. SanzidulIslam
M.Sc. In CSE, DIU
ID: 211-25-953

2. There is a huge amount of data available in the
InformationIndustry.This data is of no use until itis
converted intousefulinformation.It is necessary to
analyze this huge amount of data and extract useful
information fromit.
Extraction of information is not only the single process,
data mining also involves other processes such as Data
Cleaning, Data Integration, Data Transformation, Data
Mining, Pattern Evaluation and DataPresentation.
Once all these processes are over, we would be able touse
this information in many applications such as Fraud
detection, Market analysis, Science exploration,etc.

3. What is DataMining?
Why DataMining?
What is KDDProcess?
On What Kind ofData?
Data MiningTechniques
Data Mining QueryLanguage
Applications of DataMining

4. Extraction of interesting
Patterns or Knowledge
from huge amount ofdata
(KnowledgeDiscovery
fromData)
One of the Step fromKDD
process

5. The Explosive Growth of Data:from
terabytes topetabytes
We are drowning in data, but starvingfor
knowledge!
Fraud detection and detection of unusual
patterns

6. Datacleaning
to remove noise and inconsistentdata
Dataintegration
where multiple data sources may becombined
Dataselection
RelatedData
Datatransformation
Unifiedformat

7. Datamining
ExtractPatterns
Patternevaluation
to identify the truly interestingpatterns
representingknowledge
Knowledgepresentation
Present the mined knowledge to theuser

9. RelationalDatabases
Collection oftables
DataWarehouses
Data from differentsources
TransactionalDatabases
Consists of a file where each record represent
transactions
Advanced Data &Applications
Multimedia, Spatial data andWWW

10. Classification
Clustering
Regression
AssociationRules

11. Classification is the process of predicting theclass
of a newitem.
Therefore to classify the new item and identify to
which class itbelongs

12. Group Data into Clusters
Similar data is grouped in the samecluster
Dissimilar data is grouped in the samecluster

13. “Regression deals with the
prediction of a value, rather
than aclass.”
Regression is a data mining
function that predicts anumber
For example, a regression
model could be used topredict
children's height, given their
age, weight, and other factors.

14. “An association algorithm creates
rules that describe how often events
have occurredtogether.”
Example: When a customer buysa
Computer, then 90% of the time
they will buysoftwares.

15. A DMQL can provide the ability to supportinteractive
datamining.
Adopts SQL-likesyntax
Hence, can be easily integrated with relational query
languages

16. Market BasketAnalysis
Market basket analysis is a modeling technique based upon a theorythat
if you buy a certain group of items you are more likely to buy another
group ofitems.
This information may help the retailer to know the buyer’s needsand
retailer can enhance the store’slayout
BioInformatics
Mining biological data helps to extract useful knowledge frommassive
datasets gathered in biology, and in other related life sciencesareas
Applications of data mining to bioinformaticsinclude
gene finding, protein function inference, disease diagnosis,disease
treatment

17. Education
Data mining can be used by an institution to take accurate
decisions and also to predict the results of thestudent.
Learning pattern of the students can be captured and usedto
develop techniques to teachthem.
Customer Relationships Management(CRM)
To maintain a proper relationship with a customer a business
need to collect data and analyze theinformation.
With data mining technologies the collected data canbe
used foranalysis.