8530521.doc
Created by Chethan.M
Data Mining Goal of Data Mining Simplification and automation of the overall statistical process, from data source(s) to model application Changed over the years — Replace statistician ? Better models, less grunge work — Many different data mining algorithms / tools available — Statistical expertise required to compare different techniques — Build intelligence into the software
Data Mining Is… Decision Trees Nearest Neighbor Classification Neural Networks Rule Induction K-means Clustering
Data Mining is Not...
Data warehousing SQL / Ad Hoc Queries / Reporting Software Agents Online Analytical Processing (OLAP) Data Visualization
Why data-mining now? Data mining is an increasingly popular topic (If the number of new textbooks is anything to go by). Two main reasons: With computers now mediating most aspects of our lives, there has been a
large increase in the accumulation of electronic data. With computers being increasingly up to the demands of complex modeling, it is getting easier to process larger datasets. Why Mine Data? Commercial Viewpoint Data volumes are too large for classical analysis approaches: Large number of records High dimensional data Leverage organization’s data assets Only a small portion of the collected data is ever analyzed Data that may never be analyzed continues to be collected, at a great expense, out of fear that something which may prove important in the future is missing. Lots of data is being collected and warehoused Web data, e-commerce
ISiM
8530521.doc
Created by Chethan.M
purchases at department/grocery stores Bank/Credit Card transactions Computers have become cheaper and more powerful Competitive Pressure is Strong Provide better, customized services for an edge (e.g. In Customer Relationship Management)
Scientific Viewpoint Data collected and stored at enormous speeds (GB/hour)
remote sensors on a satellite telescopes scanning the skies micro arrays generating gene expression data scientific simulations generating terabytes of data
Traditional techniques infeasible for raw data Data mining may help scientists
In classifying and segmenting data In Hypothesis Formation Origins of Data Mining Draws ideas from machine learning/AI, pattern recognition, statistics, and database systems Traditional Techniques may be unsuitable due to Enormity of data High dimensionality of data Heterogeneous, distributed nature of data
Mining Large Data Sets - Motivation There is often information “hidden” in the data that is not readily evident Human analysts may take weeks to discover useful information Much of the data is never analyzed at all
What is Data Mining? ----- Many Definitions Data processing using sophisticated data search capabilities and statistical algorithms to discover patterns and correlations in large preexisting databases; a way to discover new meaning in data. Non-trivial extraction of implicit, previously unknown and potentially useful information from data. Exploration & analysis, by automatic or semi-automatic means, of large quantities of data in order to discover meaningful patterns.
ISiM
8530521.doc
Created by Chethan.M
The process of identifying commercially useful patterns or relationships in databases or other computer repositories through the use of advanced statistical tools. The automated extraction of predictive information from (large) databases. A step in the knowledge discovery process consisting of particular algorithms (methods) that under some acceptable objective, produces a particular enumeration of patterns (models) over the data. Data mining is the process of discovering interesting knowledge from large amounts of data stored either in databases, data warehouses, or other information repositories.
What is (not) Data Mining? What is not Data Mining? Look up phone number in phone directory Query a Web search engine for information about “Amazon”
ISiM
What is Data Mining? Certain names are more prevalent in certain US locations (O’Brien, O’Rurke, O’Reilly… in Boston area) Group together similar documents returned by search engine according to their context (e.g. Amazon rainforest, Amazon.com,)
8530521.doc
Created by Chethan.M
Statistics/AI
Machine Learning/ Pattern Recognition
Data Mining Database systems Data Mining Types: Predictive data mining: This produces the model of the system described by the given data. It uses some variables or fields in the data set to predict unknown or future values of other variables of interest. Descriptive data mining: This produces new, nontrivial information based on the available data set. It focuses on finding patterns describing the data that can be interpreted by humans.
Defining `data' By `data', we mean sets of variable values, e.g., Annual rainfall in Sussex for the last twenty years; Age, salary and IQ for all members of Sussex faculty.
Records Values are organised in combinations called records. Each record has a particular context, e.g., age, salary and IQ specifically for the Informatics HoD. Combinations may also be called vectors (esp. in neural-networks) and data-points (esp. in statistics). A single record is a datum.
ISiM
8530521.doc
Created by Chethan.M
Tabulation Data are often presented in a tabulated form, with one datum per row, and one variable per column. NAME
smith bloggs bush ...
AGE SALARY IQ
42 29 50
36K 30K 60K
130 140 120
Where data are used for prediction, the to-be-predicted variable normally appears in the final column (and is often called `class').
Basic data-types Data may be classified according to the number and character of variables involved. • • • •
Univariate/discrete: one variable with integer/symbolic values. Univariate/continuous: one variable with real/continuous values. Multivariate/discrete: more than one variable with integer/symbolic values. Multivariate/continuous: more than one variable with real/continuous values.
Data Mining Tasks...
Classification [Predictive] Clustering [Descriptive] Association Rule Discovery [Descriptive] Sequential Pattern Discovery [Descriptive] Regression [Predictive] Deviation Detection [Predictive]
ISiM
8530521.doc
Created by Chethan.M
Classification: Definition
Given a collection of records (training set ) Each record contains a set of attributes, one of the attributes is the class. Find a model for class attribute as a function of the values of other attributes. Goal: previously unseen records should be assigned a class as accurately as possible.
A test set is used to determine the accuracy of the model. Usually, the given data set is divided into training and test sets, with training set used to build the model and test set used to validate it.
Explicit and implicit structure A dataset is a body of data. Any dataset has explicit structure, ie., the numbers/values in the records. Generally, there is also implicit structure. Data mining is the task of identifying and modeling implicit structure, either as an end in itself or as a means of obtaining new information.
Example: A-level grades Dataset containing average A-level grades for the past ten years. Explicit structure is the mapping between years and average grades. (Explicit structure = `what you see')
ISiM
8530521.doc
Created by Chethan.M
There is also implicit structure---a gradual increase in values over time. (Average grades are increasing by approx 3% per year.) Classification Example
ic or
al
g
te ca
Tid Refund
ic
al
r go
te Marital Ca
us
o nu
ti
n co
s as
cl
Status
Taxable Income Cheat
1
Yes
Single
125K
No
2
No
Married
100K
No
3
No
Single
70K
No
4
Yes
Married
120K
No
5
No
Divorced 95K
Yes
6
No
Married
No
7
Yes
Divorced 220K
No
8
No
Single
85K
Yes
9
No
Married
75K
No
10
No
Single
90K
Yes
60K
Refund Marital Status
Taxable Income Cheat
No
Single
75K
?
Yes
Married
50K
?
No
Married
150K
?
Yes
Divorced 90K
?
No
Single
40K
?
No
Married
80K
?
10
Training Set
Test Set
Learn Classifier
10
Model
Challenges of Data Mining
Scalability Dimensionality Complex and Heterogeneous Data Data Quality Data Ownership and Distribution Privacy Preservation Streaming Data
ISiM
8530521.doc
Created by Chethan.M
Statistical methods Case-based reasoning Neural networks Decision trees
DM & DW: Data Warehousing + Data Mining = Increased performance of decision making process + Knowledgeable decision makers
Data Mining Applications Data Mining For Financial Data Analysis Data Mining For Telecommunications Industry Data Mining For The Retail Industry Data Mining In Healthcare and Biomedical Research Data Mining In Science and Engineering Reference:
ISiM
8530521.doc
Created by Chethan.M
1. Introduction to Data Mining by Tan, Steinbach, Kumar 2. Data Mining: Concepts and Techniques: Jiawei Han and Micheline Kamber 3. Kurt Thearling, Ph.D. An Introduction to Data Mining. www.thearling.com
Data Mining: Concepts and Techniques
ISiM