Data Mining

Ying Liu, Associate Prof., Ph.D

University of Chinese Academy of Sciences

 Review
 Data mining—core of Pattern Evaluation
knowledge discovery
process
Data Mining
Selection and
Transformation

Data Warehouse
Data Cleaning
and Integration

Databases Flat files

2015/10/13 2
Review
 Learning the application domain
 relevant prior knowledge and goals of application
 Creating a target data resource
 Data cleaning and preprocessing: (may take 60% of effort!)
 Data reduction and transformation
 Find useful features, dimensionality/variable reduction, invariant
representation
 Choosing the mining algorithm(s) to search for patterns of
interest
 Pattern evaluation and knowledge presentation
 visualization, transformation, removing redundant patterns, etc.
 Use
2015/10/13
of discovered knowledge 3
Data Preprocessing Overview

 Why preprocess the data?

 Descriptive data summarization
 Data cleaning
 Data transformation
 Data integration
 Data reduction
 Discretization and concept hierarchy generation
 Summary
2015/10/13 4
Why Data Preprocessing?

 Data in the real world is dirty

 incomplete: lacking attribute values, lacking certain
attributes of interest, or containing only aggregated
data
• e.g., occupation=“ ”
 noisy: containing errors or outliers
• e.g., Salary=“-10”
 inconsistent: containing discrepancies in codes or
names
• e.g., Age=“42” Birthday=“03/07/1997”
• e.g., Was rating “1,2,3”, now rating “A, B, C”

2015/10/13 5
Why Is Data Dirty?
 Incomplete data may come from
 “Not applicable” data value when collected
 Different considerations between the time when the data was
collected and when it is analyzed
 Human/hardware/software problems
 Noisy data (incorrect values) may come from
 Faulty data collection instruments
 Human or computer error at data entry
 Errors in data transmission
 Inconsistent data may come from
 Different data sources
 Functional dependency violation (e.g., modify some linked data)
 Duplicate records also need data cleaning
2015/10/13 6
Why Is Data Preprocessing Important?
 No quality data, no quality mining results!
 Quality decisions must be based on quality data
• e.g., duplicate or missing data may cause incorrect or even
misleading statistics

 Data extraction, cleaning, and transformation

comprises the majority of the work of building a
data warehouse

2015/10/13 7
Major Tasks in Data Preprocessing
 Data cleaning
 Fill in missing values, smooth noisy data, identify or remove
outliers, and resolve inconsistencies
 Data integration
 Integration of multiple databases, data cubes, or files
 Data transformation
 Normalization and aggregation
 Data reduction
 Obtains reduced representation in volume but produces the same
or similar analytical results
 Data discretization
 Data reduction for numerical data
2015/10/13 8
Forms of Data Preprocessing

2015/10/13 9
Data Preprocessing Overview

 Why preprocess the data?

 Descriptive data summarization
 Data cleaning
 Data integration and transformation
 Data reduction
 Discretization and concept hierarchy generation
 Summary
2015/10/13 10
Data Descriptive Characteristics
 Central tendency
 Mean, weighted mean, etc.
 Dispersion characteristics
 median, max, min, quantiles, outliers, variance, etc.
 Numerical dimensions correspond to sorted
intervals
 Data dispersion: analyzed with multiple granularities of
precision
 Boxplot or quantile analysis on sorted intervals
2015/10/13 11
Measuring the Central Tendency
 Mean (algebraic measure):
1 n
 Mean x   xi n


n i 1
Weighted arithmetic mean: w x i i
x i 1
n
 Trimmed mean: chopping extreme values w
i 1
i

 Median: A holistic measure

 Middle value if odd number of values, or average of the middle two values
otherwise

 Mode
 Value that occurs most frequently in the data
 Unimodal, bimodal, trimodal
 Empirical formula: mean  mode  3  (mean  median)
2015/10/13 12
Symmetric vs. Skewed Data
 Median, mean and mode of
symmetric, positively and
negatively skewed data

2015/10/13 13
 Measuring the Dispersion of Data
 Quartiles, outliers and boxplots
 Quartiles: Q1 (25th percentile), Q3 (75th percentile)
 Inter-quartile range: IQR = Q3 – Q1
 Five number summary: min, Q1, M, Q3, max
 Boxplot: ends of the box are the quartiles, median is marked, whiskers, and
plot outlier individually
 Outlier: usually, a value higher/lower than 1.5 x IQR

 Variance and standard deviation (sample: s, population: σ)

 Variance: (algebraic, scalable computation)
n n
1 1
1 n

1 n 2 1 n 2
[ xi  ( xi ) ]        xi   2
2 2 2
s 
2
( xi  x ) 
2 ( xi )
n  1 i 1 n  1 i 1 n i 1 N i 1 N i 1

 Standard deviation s (or σ) is the square root of variance s2 (or σ2)

2015/10/13 14
Boxplot Analysis
 Five-number summary of a distribution:
Minimum, Q1, M, Q3, Maximum
 Boxplot
 Data is represented with a box
 The ends of the box are at the first and third
quartiles, i.e., the height of the box is IQR
 The median is marked by a line within the box
 Whiskers: two lines outside the box extend to
Minimum and Maximum, terminate at 1.5xIQR

2015/10/13 15
Boxplot Analysis

2015/10/13 16
Properties of Normal Distribution
Curve
 The normal (distribution) curve
 From μ–σ to μ+σ: contains about 68% of the measurements (μ:
mean, σ: standard deviation)
 From μ–2σ to μ+2σ: contains about 95% of it
 From μ–3σ to μ+3σ: contains about 99.7% of it

2015/10/13 17
Histogram Analysis
 Graph displays of basic statistical class
descriptions
 Frequency histograms
• A univariate graphical method
• Consists of a set of rectangles that reflect the counts or
frequencies of the classes present in the given data

2015/10/13 18
Example

2015/10/13 19
Quantile Plot
 Display all of the data (allowing the user to assess both
the overall behavior and unusual occurrences)
 Plot quantile information
 For a data xi data sorted in increasing order, fi indicates that
approximately 100 fi% of the data are below or equal to the value
xi

i – 0.5
fi =
n

2015/10/13 20
Quantile-Quantile (Q-Q) Plot
 Graphs the quantiles of one univariate distribution against
the corresponding quantiles of another
 Allows the user to view whether there is a shift in going
from one distribution to another

2015/10/13 21
Scatter Plot
 Provides a first look at bivariate data to see clusters
of points, outliers, etc.
 Each pair of values is treated as a pair of coordinates
and plotted as points in the plane

2015/10/13 22
Loess Curve
 Adds a smooth curve to a scatter plot in order to provide
better perception of the pattern
 Loess curve is fitted by setting two parameters: a
smoothing parameter, and the degree of the
polynomials that are fitted by the regression

2015/10/13 23
Graphic Displays of Basic Statistical
Descriptions
 Boxplot
 Histogram
 Quantile plot: each value xi is paired with fi
indicating that approximately 100 fi % of data are  xi
 Quantile-quantile (q-q) plot: graphs the quantiles of
one univariant distribution against the corresponding
quantiles of another
 Scatter plot: each pair of values is a pair of
coordinates and plotted as points in the plane
 Loess (local regression) curve: add a smooth curve
to a scatter plot to provide better perception of the
pattern of dependence

2015/10/13 24
Exercise

1. The values of data tuples are 13, 15, 16, 16,

19, 20, 20, 21.
(a) What is the mean of the data? What is the
median?
(b) What is the mode of the data?
(c) What is Q1 and Q3?
(d) What is the IQR of the data?
(e) Give the five-number-summary of the data.
(f) Show a boxplot for the data.
2015/10/13 25
 Overview: Data Preprocessing

 Why preprocess the data?

 Descriptive data summarization
 Data cleaning
 Data transformation
 Data integration
 Data reduction
 Discretization and concept hierarchy generation
 Summary
2015/10/13 26
Data Cleaning

 Importance
 “Data cleaning is one of the three biggest problems
in data warehousing”—Ralph Kimball
 “Data cleaning is the number one problem in data
warehousing”—DCI survey

 Data cleaning tasks

 Fill in missing values
 Identify outliers and smooth out noisy data
 Correct inconsistent data
 Resolve redundancy caused by data integration
2015/10/13 27
Missing Data
 Data is not always available
 E.g., many tuples have no recorded value for several attributes,
such as customer income in sales data
 Missing data may be due to
 equipment malfunction
 inconsistent with other recorded data and thus deleted
 data not entered due to misunderstanding
 certain data may not be considered important at the time of
entry

2015/10/13 28
How to Handle Missing Data?
 Ignore the tuple: usually done when class label is missing
(assuming the tasks in classification—not effective when
the percentage of missing values per attribute varies
considerably).
 Fill in the missing value manually: tedious + infeasible?
 Fill in it automatically with
 a global constant : e.g., “unknown”, a new class?!
 the attribute mean
 the attribute mean for all samples belonging to the same class:
smarter
 the most probable value: inference-based such as Bayesian
formula or decision tree
2015/10/13 29
Noisy Data

 Noise: random error or variance in a measured variable

 Incorrect attribute values may due to
 faulty data collection instruments
 data entry problems
 data transmission problems
 technology limitation
 inconsistency in naming convention
 Other data problems which requires data cleaning
 duplicate records
 inconsistent data

2015/10/13 30
How to Handle Noisy Data?
 Binning
 first sort data and partition into bins
 Smooth noise by consulting its neighbors, local smooth
 then one can smooth by bin means, smooth by bin
median, smooth by bin boundaries, etc.
 Regression
 smooth by fitting the data into regression functions
 Clustering
 detect and remove outliers
 Combined computer and human inspection
 detect suspicious values and check by human (e.g.,
deal with possible outliers)
2015/10/13 31
Simple Discretization Methods: Binning
 Equal-width (distance) partitioning
 Divides the range into N intervals of equal size
 if A and B are the lowest and highest values of the attribute, the
width of intervals will be: W = (B –A)/N
 The most straightforward, but outliers may dominate presentation
 Skewed data is not handled well

 Equal-depth (frequency) partitioning

 Divides the range into N intervals, each containing approximately
same number of samples
 Good data scaling

2015/10/13 32
Binning Methods for Data Smoothing
 Sorted data for price (in dollars): 4, 8, 9, 15, 21, 21, 24, 25, 26, 28,
29, 34
* Partition into equal-frequency (equi-depth) bins:
- Bin 1: 4, 8, 9, 15
- Bin 2: 21, 21, 24, 25
- Bin 3: 26, 28, 29, 34
* Smoothing by bin means:
- Bin 1: 9, 9, 9, 9
- Bin 2: 23, 23, 23, 23
- Bin 3: 29, 29, 29, 29
* Smoothing by bin boundaries:
- Bin 1: 4, 4, 4, 15
- Bin 2: 21, 21, 25, 25
- Bin 3: 26, 26, 26, 34

2015/10/13 33
Regression
y

Y1

Y1* y=x+1

X1 x

2015/10/13 34
Cluster Analysis

2015/10/13 35
Exercise
1. Suppose a group of 12 sales price records has been
sorted as follows: 5, 10, 11, 13, 15, 15, 15, 55, 60, 60,
65, 65.
(a) Smooth the data by bin means, using a bin depth of 4.
(b) Smooth the data by bin boundaries, using a bin depth
of 4.
(c) Smooth the data by bin means, using 3 bins of equal-
width partitioning.

2015/10/13 36
Overview: Data Preprocessing

 Why preprocess the data?

 Descriptive data summarization
 Data cleaning
 Data transformation
 Data integration
 Data reduction
 Discretization and concept hierarchy generation
 Summary
2015/10/13 37
Data Transformation
 Smoothing: remove noise from data
 Binning, regression, clustering
 Normalization: scaled to fall within a small, specified
range
 min-max normalization
 z-score normalization
 normalization by decimal scaling
 Aggregation: summarization, data cube construction
 Attribute/feature construction
 New attributes constructed from the given ones

2015/10/13 38
Data Transformation: Normalization
 Min-max normalization: to [new_minA, new_maxA]
v  minA
v'  (new _ maxA  new _ minA)  new _ minA
maxA  minA
 Ex. Let income range $12,000 to $98,000 normalized to [0.0, 1.0].
73,600  12,000
Then $73,600 is mapped to 98,000  12,000
(1.0  0)  0  0.716

 Z-score normalization (μ: mean, σ: standard deviation):

v  A 73,600  54,000
v'   1.225
 A
16,000
 Ex. Let μ = 54,000, σ = 16,000. Then
 Normalization by decimal scaling
v
v'  j Where j is the smallest integer such that Max(|ν’|) < 1
10
 Ex. (-986, 917) => (-0.986, 0.917), j=3
2015/10/13 39
Exercise
1. Use two methods to normalize the following
group of data: 200, 300, 400, 600, 1000.
(a) min-max normalization by setting min=0 and
max = 1.
(b) Z-score normalization (mean = 500, standard
deviation = 316.2).

2015/10/13 40
Overview: Data Preprocessing

 Why preprocess the data?

 Descriptive data summarization
 Data cleaning
 Data transformation
 Data integration
 Data reduction
 Discretization and concept hierarchy generation
 Summary
2015/10/13 41
Data Integration
 Data integration:
 Combines data from multiple sources into a coherent store
 Schema integration: e.g., A.cust-id  B.cust-#
 Integrate metadata from different sources
 Entity identification problem:
 Identify real world entities from multiple data sources, e.g., Bill
Clinton = William Clinton
 Detecting and resolving data value conflicts
 For the same real world entity, attribute values from different
sources are different
 Possible reasons: different representations, different scales, e.g.,
metric vs. British units

2015/10/13 42
Handling Redundancy in Data Integration

 Redundant data often occur

 Derivable data: One attribute may be a “derived”
attribute in another table, e.g., payment, payment rate
 Redundant attributes may be able to be detected
by correlation analysis

2015/10/13 43
Correlation Analysis (Numerical Data)
 Correlation coefficient (also called Pearson’s product
moment coefficient)

rA ,B 
 (ai  A )(bi  B )

 (a b )  n A B
i i

(n  1)AB (n  1)AB

where n is the number of tuples, A and B are the respective means

of A and B, σA and σB are the respective standard deviation of A
and B, and Σ(aibi) is the dot-product of A and B.
 If rA,B > 0, A and B are positively correlated (A’s values
increase as B’s). The higher, the stronger correlation.
 rA,B = 0: independent
 rA,B
2015/10/13 < 0: negatively correlated 44
Positively and Negatively Correlated Data

(a) (b)

2015/10/13 45
(c)
 Not Correlated Data

2015/10/13 46
Correlation Analysis (Categorical Data)
 Χ2 (chi-square) test
(Observed  Expected ) 2
 
2

Expected
 The larger the Χ2 value, the more likely the variables
are correlated
 The cells that contribute the most to the Χ2 value are
those whose actual count is very different from the
expected count
 Correlation does not imply causality
 # of hospitals and # of car-theft in a city are correlated
 Both are causally linked to the third variable: population

2015/10/13 47
Chi-Square Calculation: An Example
Play chess Not play chess Sum (row)
Like science fiction 250(90) 200(360) 450
Not like science fiction 50(210) 1000(840) 1050
Sum(col.) 300 1200 1500

 Χ2 (chi-square) calculation (numbers in parenthesis are

expected counts calculated based on the data distribution
in the two categories)
(250  90) 2 (50  210) 2 (200  360) 2 (1000  840) 2
 2
    507.93
90 210 360 840

 It shows that like_science_fiction and play_chess are

correlated in the group, dependent
2015/10/13 48
Handling Redundancy in Data Integration

 Careful integration of the data from multiple

sources may help reduce/avoid redundancies
and inconsistencies and improve mining speed
and quality

2015/10/13 49
Excerise
1. The following contingence table summarizes
supermarket transaction data.
(a) Based on the given data, is the purchase of hot dogs
independent of the purchase of hamburgers?
(b) If correlated, what kind of correlation relationship
exists between the two items?
hot dogs not hot dogs sum
hamburgers 4000 3500 7500
not hamburgers 2000 500 2500
sum 6000 4000 10000

2015/10/13 50
Overview: Data Preprocessing

 Why preprocess the data?

 Descriptive data summarization
 Data cleaning
 Data transformation
 Data integration
 Data reduction
 Discretization and concept hierarchy generation
 Summary

2015/10/13 51
Data Reduction

 Why data reduction?

 A database/data warehouse may store terabytes of data

 Complex data analysis/mining may take a very long time to run

on the complete data set

 Data reduction
 Obtain a reduced representation of the data set that is much
smaller in volume but yet produce the same (or almost the same)
analytical results

2015/10/13 52
Data Reduction Strategies

 Data reduction strategies

 Data cube aggregation

 Dimensionality reduction — e.g., remove unimportant attributes

 Data compression

 Numerosity reduction — e.g., fit data into models

 Discretization and concept hierarchy generation

2015/10/13 53
Data Cube Aggregation
 The lowest level of a data cube (base cuboid)
 The aggregated data for an individual entity of interest
 Multiple levels of aggregation in data cubes
 Further reduce the size of data to deal with
 Reference appropriate concept levels
 Use the smallest representation which is enough to
solve the task

2015/10/13 54
Attribute Subset Selection

 Feature selection (i.e., attribute subset selection):

 Select a minimum set of features such that the probability
distribution of different classes given the values for those
features is as close as possible to the original distribution
given the values of all features
 reduce # of features in the patterns, easier to understand
 Heuristic methods (due to exponential # of choices):
 Step-wise forward selection
 Step-wise backward elimination
 Combining forward selection and backward elimination
 Decision-tree induction

2015/10/13 55
Feature Selection Methods
 There are 2d possible sub-features of d features
 Greedy methods: locally optimal
 Choose by “statistical significance” tests
 Best step-wise forward selection:
• The best single-feature is picked first
• Then next best feature condition to the first, ...
 Step-wise backward elimination:
• Repeatedly eliminate the worst feature
 Best combined feature selection and elimination
 Decision tree induction

2015/10/13 56
Example

2015/10/13 57
Data Compression
 String compression
 There are extensive theories and well-tuned
algorithms
 Typically lossless
 But only limited manipulation is possible without
expansion
 Audio/video compression
 Typically lossy compression, with progressive
refinement
 Sometimes small fragments of signal can be
reconstructed without reconstructing the whole
2015/10/13 58
Data Compression

Original Data Compressed

Data
lossless

Original Data
Approximated

2015/10/13 59
Wavelet Transformation

 Discrete wavelet transform (DWT): linear signal

processing, multi-resolutional analysis
 Compressed approximation: store only a small fraction of
the strongest wavelet coefficients
 Similar to discrete Fourier transform (DFT), but better
lossy compression, localized in space
 Method:
 Length, L, must be an integer power of 2 (padding with 0’s, when
necessary)
 Each transform has 2 functions: smoothing, difference
 Applies to pairs of data, resulting in two sets of data of length L/2
 Applies two functions recursively, until reaches the desired length
2015/10/13 60
Dimensionality Reduction: Principal
Component Analysis (PCA)
 Given N data vectors from n-dimensions, find k ≤ n vectors
(principal components) that can be best used to represent data
 Steps
 Normalize input data: Each attribute falls within the same range
 Compute k vectors, i.e., principal components
 Each input data (vector) is a linear combination of the k principal
component vectors
 The principal components are sorted in order of decreasing
“significance” or strength
 Since the components are sorted, the size of the data can be
reduced by eliminating the weak components, i.e., those with low
variance. (i.e., using the strongest principal components, it is
possible to reconstruct a good approximation of the original data)
 Work
2015/10/13
for numeric data only 61
Principal Component Analysis
X2

Y1
Y2

X1

2015/10/13 62
Numerosity Reduction
 Reduce data volume by choosing alternative,
smaller forms of data representation
 Parametric methods
 Assume the data fits some model, estimate model
parameters, store only the parameters, and discard the
data (except possible outliers)

 Non-parametric methods
 Do not assume models
 Major families: histograms, clustering, sampling

2015/10/13 63
Data Reduction Method (1):
Regression Models
 Linear regression: Data are modeled to fit a
straight line
 Often uses the least-square method to fit the line

 Y =  + X，two regression coefficients

 Multiple regression: allows a response variable Y
to be modeled as a linear function of
multidimensional feature vector
 Y =  + 1X1 + 2X2

2015/10/13 64
 Data Reduction Method (2): Histograms
 Divide data into buckets and
store frequency for each
bucket
 Partitioning rules:
 Equal-width: equal bucket range
 Equal-frequency (or equal-depth)
 V-optimal: with the least
histogram variance (weighted
sum of the original values that
each bucket represents)

2015/10/13 65
Data Reduction Method (3): Clustering

 Partition data set into clusters based on

similarity
 Store cluster representation (e.g., centroid
and diameter) only
 Can be very effective for data that can be
clustered but not for “smeared” data
 There are many choices of clustering
definitions and clustering algorithms

2015/10/13 66
Data Reduction Method (4): Sampling
 Sampling: obtaining a small sample s to
represent the whole data set N
 Choose a representative subset of the data
 Simple random sampling may have very poor
performance in the presence of skew
 Adaptive sampling methods
 Stratified sampling:
• Approximate the percentage of each class (or
subpopulation of interest) in the overall database
• Used in conjunction with skewed data
 Fast, scan database once
2015/10/13 67
Sampling:With or Without Replacement

Raw Data
2015/10/13 68
Sampling: Cluster or Stratified Sampling

Raw Data Cluster/Stratified Sample

2015/10/13 69
Overview: Data Preprocessing
 Why preprocess the data?
 Descriptive data summarization
 Data cleaning
 Data transformation
 Data integration
 Data reduction
 Discretization and concept hierarchy generation
 Summary
2015/10/13 70
Discretization and Concept Hierarchy
 Discretization
 Reduce the number of values for a given continuous
attribute by dividing the range of the attribute into
intervals
 Interval labels can then be used to replace actual data
values
 Discretization can be performed recursively on an
attribute
 Concept hierarchy formation
 Recursively reduce the data by collecting and replacing
low level concepts (such as numeric values for age) by
2015/10/13
higher level concepts (such as young, middle-aged,
71
or
 Example of Concept Hierarchy
$0…$1000

$0…$200 $200…$400 $400…$600 $600…$800 $800…$1000

$0…$100 $200…$300 $400…$500 $600…$700 $800…$900

$100…$200 $300…$400 $500…$600 $700…$800 $900…$1000

2015/10/13 72
Discretization and Concept Hierarchy
Generation for Numeric Data
 Typical methods: All the methods can be applied
recursively
 Binning
• Top-down split, replace the value by bin mean or median

 Histogram analysis
• Top-down split

 Clustering analysis
• Top-down split

 Entropy-based discretization: supervised, top-down split

 Segmentation by natural partitioning: top-down split

2015/10/13 73
Entropy-Based Discretization
 Entropy is calculated based on class distribution of the samples in the
set. Given m classes, the entropy of S is
m
Entropy(S )   p
i 1
i log2( pi ) where pi is the probability of class i in S
 Given a set of samples S, if S is partitioned into two intervals S1 and
S2 using boundary T, the entropy after partitioning is
| S1 | | |
Entropy(S ,T )  Entropy(S 1)  S 2 Entropy(S 2)
|S | |S |
 The boundary that maximizes the information gain over all possible
boundaries is selected as a binary discretization
Gain(S ,T )  Entropy(S )  Entropy(S ,T )

 The process is recursively applied to partitions obtained until some

stopping criterion is met
 Such a boundary may reduce data size and improve classification
accuracy
2015/10/13 74
An Example of Entropy-based
Partitioning
age income student credit_rating buys_computer
<=30 high no fair no
<=30 high no excellent no
31…40 high no fair yes
>40 medium no fair yes
>40 low yes fair yes
>40 low yes excellent no
31…40 low yes excellent yes
<=30 medium no fair no
<=30 low yes fair yes
>40 medium yes fair yes
<=30 medium yes excellent yes
31…40 medium no excellent yes
31…40 high yes fair yes
>40 medium no excellent no

2015/10/13 75
An Example of Entropy-based
Partitioning
Gain(age )  0.246
Gain(income )  0.029
Gain(student )  0.151
Gain(credit _ rating )  0.048
<=30 30..40 >40

2015/10/13 76
Segmentation by Natural Partitioning
 A simply 3-4-5 rule can be used to segment numeric data
into relatively uniform, “natural” intervals
 If an interval covers 3, 6, 7 or 9 distinct values at the “most
significant digit”, partition the range into 3 equal-width intervals
 If it covers 2, 4, or 8 distinct values at the “most significant digit”,
partition the range into 4 intervals
 If it covers 1, 5, or 10 distinct values at the “most significant digit”,
partition the range into 5 intervals
 The top-level segmentation represents the majority

2015/10/13 77
 Example of 3-4-5 Rule
count

Step 1: -$351 -$159 profit $1,838 $4,700

Min Low (i.e, 5%-tile) High(i.e, 95%-0 tile) Max
Step 2: msd=1,000 Low=-$1,000 High=$2,000

(-$1,000 - $2,000)
Step 3:

(-$1,000 - 0) (0 -$ 1,000) ($1,000 - $2,000)

(-$400 -$5,000)
Step 4:

(-$400 - 0) ($2,000 - $5, 000)

(0 - $1,000) ($1,000 - $2, 000)
(0 -
(-$400 - ($1,000 -
$200)
$1,200) ($2,000 -
-$300)
($200 - $3,000)
($1,200 -
(-$300 - $400)
$1,400)
-$200) ($3,000 -
($400 - ($1,400 - $4,000)
(-$200 - $600) $1,600) ($4,000 -
-$100) $5,000)
($600 - ($1,600 -
$800) ($800 - ($1,800 -
$1,800)
(-$100 - $1,000) $2,000)
0)
2015/10/13 78
Concept Hierarchy Generation for
Categorical Data
 Specification of a partial/total ordering of attributes
explicitly at the schema level by users or experts
 street < city < state < country
 Specification of a hierarchy for a set of values by
explicit data grouping
 {Urbana, Champaign, Chicago} < llinois
 Automatic generation of hierarchies (or attribute
levels) by the analysis of the number of distinct
values
 E.g., for a set of attributes: {street, city, state, country}

2015/10/13 79
Automatic Concept Hierarchy
Generation
 Some hierarchies can be automatically generated
based on the analysis of the number of distinct values
per attribute in the data set
 The attribute with the most distinct values is placed at the
lowest level of the hierarchy
 Exceptions, e.g., weekday, month, quarter, year
country 15 distinct values

province_or_ state 365 distinct values

city 3567 distinct values

street 674,339 distinct values

2015/10/13 80
Overview: Data Preprocessing

 Why preprocess the data?

 Descriptive data summarization
 Data cleaning
 Data transformation
 Data integration
 Data reduction
 Discretization and concept hierarchy generation
 Summary
2015/10/13 81
Summary
 Data preparation or preprocessing is a big issue for
both data warehousing and data mining
 Descriptive data summarization is needed for quality
data preprocessing
 Data preparation includes
 Data cleaning and data integration
 Data reduction and feature selection
 Discretization
 A lot a methods have been developed but data
preprocessing still an active area of research

2015/10/13 82