Data Preprocessing

(based on notes by Jiawei Han and Micheline Kamber)

 Agenda
• Why data preprocessing?
• Data cleaning
• Data integration and transformation
• Data reduction
• Discretization and concept hierarchy generation
• Summary
 Why Data Preprocessing?
• Data in the real world is dirty
– incomplete: lacking attribute values, lacking certain attributes
of interest, or containing only aggregate data
– noisy: containing errors or outliers
– inconsistent: containing discrepancies in codes or names
• No quality data, no quality mining results!
– Quality decisions must be based on quality data
– Data warehouse needs consistent integration of quality data
• A multi-dimensional measure of data quality:
– A well-accepted multi-dimensional view:
• accuracy, completeness, consistency, timeliness, believability, value
added, interpretability, accessibility
– Broad categories:
• intrinsic, contextual, representational, and accessibility.
 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, files, or notes
• Data transformation
– Normalization (scaling to a specific range)
– Aggregation
• Data reduction
– Obtains reduced representation in volume but produces the same or similar
analytical results
– Data discretization: with particular importance, especially for numerical data
– Data aggregation, dimensionality reduction, data compression,generalization
Forms of data preprocessing
 Agenda

• Why preprocess the data?

• Data cleaning
• Data integration and transformation
• Data reduction
• Discretization and concept hierarchy generation
• Summary
 Data Cleaning

• Data cleaning tasks

– Fill in missing values
– Identify outliers and smooth out noisy data
– Correct inconsistent data
 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
– not register history or changes of the data
• Missing data may need to be inferred
 How to Handle Missing Data?
• Ignore the tuple: usually done when class label is missing (assuming the
task is classification—not effective in certain cases)

• Fill in the missing value manually: tedious + infeasible?

• Use a global constant to fill in the missing value: e.g.,
“unknown”, a new class?!

• Use the attribute mean to fill in the missing value

• Use the attribute mean for all samples of the same class to fill
in the missing value: smarter
• Use the most probable value to fill in the missing value:
inference-based such as regression, Bayesian formula, decision tree
 Noisy Data
• Q: What is noise?
• A: Random error in a measured variable.
• Incorrect attribute values may be 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
– incomplete data
– inconsistent data
 How to Handle Noisy Data?
• Binning method:
– first sort data and partition into (equi-depth) bins
– then one can smooth by bin means, smooth by bin median,
smooth by bin boundaries, etc.
– used also for discretization (discussed later)
• Clustering
– detect and remove outliers
• Semi-automated method: combined computer and
human inspection
– detect suspicious values and check manually
• Regression
– smooth by fitting the data into regression functions
Simple Discretization Methods: Binning
• Equal-width (distance) partitioning:
– It divides the range into N intervals of equal size: uniform
grid
– 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:
– It divides the range into N intervals, each containing
approximately same number of samples
– Good data scaling
– Managing categorical attributes can be tricky.
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 (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
Cluster Analysis
 Regression
y

Y1

Y1’ y=x+1

X1 x
•Linear regression (best line to fit
two variables)
•Multiple linear regression (more
than two variables, fit to a
multidimensional surface
 How to Handle Inconsistent Data?
• Manual correction using external references
• Semi-automatic using various tools
– To detect violation of known functional
dependencies and data constraints
– To correct redundant data
 Agenda

• Why preprocess the data?

• Data cleaning
• Data integration and transformation
• Data reduction
• Discretization and concept hierarchy generation
• Summary
 Data Integration
• Data integration:
– combines data from multiple sources into a coherent store
• Schema integration
– integrate metadata from different sources
– Entity identification problem: identify real world entities
from multiple data sources, e.g., A.cust-id  B.cust-#
• 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, different currency
 Handling Redundant Data in
Data Integration
• Redundant data occur often when integrating multiple DBs
– The same attribute may have different names in different databases
– One attribute may be a “derived” attribute in another table, e.g.,
annual revenue
• Redundant data may be able to be detected by correlational
analysis
( A  A)( B  B )
rA, B 
(n  1) A B
• Careful integration can help reduce/avoid redundancies and
inconsistencies and improve mining speed and quality
 Data Transformation
• Smoothing: remove noise from data (binning,
clustering, regression)
• Aggregation: summarization, data cube construction
• Generalization: concept hierarchy climbing
• Normalization: scaled to fall within a small,
specified range
– min-max normalization
– z-score normalization
– normalization by decimal scaling
• Attribute/feature construction
– New attributes constructed from the given ones
Data Transformation: Normalization
Particularly useful for classification (NNs, distance measurements,
nn classification, etc)
• min-max normalization
v  min A
v'  (new _ max  new _ min )  new _ min
A A A
max  min
A A

• z-score normalization
v  mean A
v' 
stand _ dev A

• normalization by decimal scaling

v
v'  j Where j is the smallest integer such that Max(| v ' |)<1
10
 Agenda
• Why preprocess the data?
• Data cleaning
• Data integration and transformation
• Data reduction
• Discretization and concept hierarchy generation
• Summary
 Data Reduction
• Problem:
Data Warehouse may store terabytes of data:
Complex data analysis/mining may take a very
long time to run on the complete data set

• Solution?
– Data reduction…
 Data Reduction
•Obtains a reduced representation of the data
set that is much smaller in volume but yet
produces the same (or almost the same)
analytical results
•Data reduction strategies
–Data cube aggregation
–Dimensionality reduction
–Data compression
–Numerosity reduction
–Discretization and concept hierarchy generation
 Data Cube Aggregation
• Multiple levels of aggregation in data cubes
– Further reduce the size of data to deal with
• Reference appropriate levels
– Use the smallest representation capable to solve the
task
• Queries regarding aggregated information should
be answered using data cube, when possible
 Dimensionality Reduction
• Problem: 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
– Nice side-effect: reduces # of attributes in the discovered patterns
(which are now easier to understand)
• Solution: Heuristic methods (due to exponential # of
choices) usually greedy:
– step-wise forward selection
– step-wise backward elimination
– combining forward selection and backward elimination
– decision-tree induction
 Example of Decision Tree Induction
nonleaf nodes: tests
branches: outcomes of tests
leaf nodes: class prediction
Initial attribute set:
A4 ? {A1, A2, A3, A4, A5, A6}

A1? A6?

Class 1 Class 2 Class 1 Class 2

> Reduced attribute set: {A1, A4, A6}

 Data Compression
• String compression
– There are extensive theories and well-tuned algorithms
– Typically lossless
– But only limited manipulation is possible without expansion
• Audio/video, image compression
– Typically lossy compression, with progressive refinement
– Sometimes small fragments of signal can be reconstructed
without reconstructing the whole
• Time sequence is not audio
– Typically short and vary slowly with time
 Data Compression

Original Data Compressed

Data
lossless

os sy
l
Original Data
Approximated
 Wavelet Transforms
• Discrete wavelet transform (DWT):
linear signal processing Haar2 Daubechie4

• Compressed approximation: store only a small fraction of

the strongest of the wavelet coefficients
• Similar to discrete Fourier transform (DFT), but better lossy
compression, localized in space (conserves local details)
• Method (hierarchical pyramid algorithm):
– Length, L, must be an integer power of 2 (padding with 0s, when necessary)
– Each transform has 2 functions:
• smoothing (e.g., sum, weighted avg.), weighted difference
– Applies to pairs of data, resulting in two sets of data of length L/2
– Applies the two functions recursively, until reaches the desired length
 Principal Component Analysis (PCA)
Karhunen-Loeve (K-L) method
• Given N data vectors from k-dimensions, find
c <= k orthogonal vectors that can be best used
to represent data
– The original data set is reduced (projected) to one
consisting of N data vectors on c principal components
(reduced dimensions)
• Each data vector is a linear combination of the c
principal component vectors
• Works for ordered and unordered attributes
• Used when the number of dimensions is large
 Principal Component Analysis
•The principal components (new set of axes) give important information about
variance.
•Using the strongest components one can reconstruct
X2 a good approximation of the
original signal.
Y1
Y2

X1
 Numerosity Reduction

• Parametric methods
– Assume the data fits some model, estimate model
parameters, store only the parameters, and discard the
data (except possible outliers)
– E.g.: Log-linear models: obtain value at a point in m-D
space as the product on appropriate marginal subspaces
• Non-parametric methods
– Do not assume models
– Major families: histograms, clustering, sampling
Regression and Log-Linear Models
• Linear regression: Data are modeled to fit a straight
line:
– Often uses the least-square method to fit the line
• Multiple regression: allows a response variable y to
be modeled as a linear function of multidimensional
feature vector (predictor variables)
• Log-linear model: approximates discrete
multidimensional joint probability distributions
Regression Analysis and Log-Linear Models

• Linear regression: Y =  +  X
– Two parameters ,  and  specify the line and are to be
estimated by using the data at hand.
– using the least squares criterion to the known values of Y1,
Y2, …, X1, X2, ….
• Multiple regression: Y = b0 + b1 X1 + b2 X2.
– Many nonlinear functions can be transformed into the above.
• Log-linear models:
– The multi-way table of joint probabilities is approximated
by a product of lower-order tables.
– Probability: p(a, b, c, d) = ab acad bcd
 Histograms
• Approximate data
distributions
• Divide data into buckets and
store average (sum) for each
bucket
• A bucket represents an
attribute-value/frequency
pair
• Can be constructed
optimally in one dimension
using dynamic programming
• Related to quantization
problems.
 Clustering
• Partition data set into clusters, and store cluster representation only
• Quality of clusters measured by their diameter (max distance
between any two objects in the cluster) or centroid distance (avg.
distance of each cluster object from its centroid)
• Can be very effective if data is clustered but not if data is “smeared”
• Can have hierarchical clustering (possibly stored in multi-
dimensional index tree structures (B+-tree, R-tree, quad-tree, etc))
• There are many choices of clustering definitions and clustering
algorithms (further details later)
 Sampling
• Allow a mining algorithm to run in complexity that is
potentially sub-linear to the size of the data
• Cost of sampling: proportional to the size of the sample,
increases linearly with the number of dimensions
• Choose a representative subset of the data
– Simple random sampling may have very poor performance in the
presence of skew
• Develop 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
• Sampling may not reduce database I/Os (page at a time).
• Sampling: natural choice for progressive refinement of a
reduced data set.
 Sampling

W O R
SRS le random
i m p ho ut
( s e wi t
l
samp ment)
p l a ce
re

SRSW
R

Raw Data
 Sampling
Raw Data Cluster/Stratified Sample
 Hierarchical Reduction
• Use multi-resolution structure with different degrees of
reduction
• Hierarchical clustering is often performed but tends to
define partitions of data sets rather than “clusters”
• Parametric methods are usually not amenable to
hierarchical representation
• Hierarchical aggregation
– An index tree hierarchically divides a data set into partitions
by value range of some attributes
– Each partition can be considered as a bucket
– Thus an index tree with aggregates stored at each node is a
hierarchical histogram
 Agenda

• Why preprocess the data?

• Data cleaning
• Data integration and transformation
• Data reduction
• Discretization and concept hierarchy generation
• Summary
 Discretization/Quantization
• Three types of attributes:
– Nominal — values from an unordered set
– Ordinal — values from an ordered set
– Continuous — real numbers
• Discretization/Quantization:
 divide the range of a continuous attribute into intervals
x1 x2 x3 x4 x5

y1 y2 y3 y4 y5 y6
– Some classification algorithms only accept categorical
attributes.
– Reduce data size by discretization
– Prepare for further analysis
 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.
• Concept Hierarchies
– reduce the data by collecting and replacing low level
concepts (such as numeric values for the attribute age) by
higher level concepts (such as young, middle-aged, or
senior).
Discretization and concept hierarchy
generation for numeric data
• Hierarchical and recursive decomposition using:

– Binning (data smoothing)

– Histogram analysis (numerosity reduction)

– Clustering analysis (numerosity reduction)

• Entropy-based discretization

• Segmentation by natural partitioning

 Entropy-Based Discretization
• Given a set of samples S, if S is partitioned into two intervals S1
and S2 using threshold T on the value of attribute A, the information
gain resulting from the partitioning is:
| S1 | | S2 |
I (S , T )  E ( S 1)  E ( S 2)
|S| |S|
where the entropy function E for a given set is calculated based on
the class distribution of the samples in the set. Given m classes the
m
entropy of S1 is:
E ( S1 )   pi log 2 ( pi )
i 1

where pi is the probability of class i in S1.

• The threshold that maximizes the information gain over all possible
thresholds is selected as a binary discretization.
• The process is recursively applied to partitions obtained until some
stopping criterion is met, e.g., E ( S )  I ( S , T )  
• Experiments show that it may reduce data size and improve
classification accuracy
 Segmentation by natural partitioning
• 3-4-5 rule can be used to segment numeric data into
relatively uniform, “natural” intervals.
• It partitions a given range into 3,4, or 5 equiwidth
intervals recursively level-by-level based on the value
range of the most significant digit.
* If an interval covers 3, 6, 7 or 9 distinct values at the most
significant digit, partition the range into 3 equi-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
 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)

(-$4000 -$5,000)
Step 4:

($2,000 - $5, 000)

(-$400 - 0) (0 - $1,000) ($1,000 - $2, 000)
(0 -
($1,000 -
(-$400 - $200)
$1,200) ($2,000 -
-$300) $3,000)
($200 -
($1,200 -
$400)
(-$300 - $1,400)
($3,000 -
-$200)
($400 - ($1,400 - $4,000)
(-$200 - $600) $1,600) ($4,000 -
-$100) ($600 - ($1,600 - $5,000)
$800) ($800 - ($1,800 -
$1,800)
(-$100 - $1,000) $2,000)
0)
Concept hierarchy generation for
categorical data
• Categorical data: no ordering among values
• Specification of a partial ordering of attributes
explicitly at the schema level by users or experts
• Specification of a portion of a hierarchy by
explicit data grouping
• Specification of a set of attributes, but not of their
partial ordering
• Specification of only a partial set of attributes
 Concept hierarchy generation w/o data
semantics - Specification of a set of attributes
Concept hierarchy can be automatically generated
based on the number of distinct values per attribute
in the given attribute set. The attribute with the most
distinct values is placed at the lowest level of the
hierarchy (limitations?)

country 15 distinct values

province_or_ state 65 distinct

values
city 3567 distinct values

street 674,339 distinct values

 Agenda
• Why preprocess the data?
• Data cleaning
• Data integration and transformation
• Data reduction
• Discretization and concept hierarchy generation
• Summary
 Summary
• Data preparation is a big issue for both warehousing
and mining
• Data preparation includes
– Data cleaning and data integration
– Data reduction and feature selection
– Discretization

• A lot a methods have been developed but still an

active area of research