Classification: Basic Concepts

 Classification: Basic Concepts

 Decision Tree Induction
 Bayes Classification Methods
 Rule-Based Classification
 Model Evaluation and Selection
 Techniques to Improve Classification Accuracy:
Ensemble Methods
 Summary
1
 Supervised vs. Unsupervised Learning

 Supervised learning (classification)

 Supervision: The training data (observations,
measurements, etc.) are accompanied by labels indicating
the class of the observations
 New data is classified based on the training set
 Unsupervised learning (clustering)
 The class labels of training data is unknown
 Given a set of measurements, observations, etc. with the
aim of establishing the existence of classes or clusters in
the data
2
 Prediction Problems: Classification vs.
Numeric Prediction
 Classification
 predicts categorical class labels (discrete or nominal)

 classifies data (constructs a model) based on the training

set and the values (class labels) in a classifying attribute

and uses it in classifying new data
 Numeric Prediction
 models continuous-valued functions, i.e., predicts unknown

or missing values
 Typical applications
 Credit/loan approval:

 Medical diagnosis: if a tumor is cancerous or benign

 Fraud detection: if a transaction is fraudulent

 Web page categorization: which category it is

3
 Classification—A Two-Step Process
 Model construction: describing a set of predetermined classes
 Each tuple/sample is assumed to belong to a predefined class, as

determined by the class label attribute

 The set of tuples used for model construction is training set

 The model is represented as classification rules, decision trees, or

mathematical formulae
 Model usage: for classifying future or unknown objects
 Estimate accuracy of the model

 The known label of test sample is compared with the classified

result from the model

 Accuracy rate is the percentage of test set samples that are

correctly classified by the model

 Test set is independent of training set (otherwise overfitting)

 If the accuracy is acceptable, use the model to classify new data

 Note: If the test set is used to select models, it is called validation (test) set
4
 Process (1): Model Construction

Classification
Algorithms
Training
Data

NAME RANK YEARS TENURED Classifier

Mike Assistant Prof 3 no (Model)
Mary Assistant Prof 7 yes
Bill Professor 2 yes
Jim Associate Prof 7 yes IF rank = ‘professor’
Dave Assistant Prof 6 no
OR years > 6
Anne Associate Prof 3 no
THEN tenured = ‘yes’
5
Process (2): Using the Model in Prediction

Classifier

Testing
Data Unseen Data

(Jeff, Professor, 4)
NAME RANK YEARS TENURED
Tom Assistant Prof 2 no Tenured?
Merlisa Associate Prof 7 no
George Professor 5 yes
Joseph Assistant Prof 7 yes
6
Chapter 8. Classification: Basic Concepts

 Classification: Basic Concepts

 Decision Tree Induction
 Bayes Classification Methods
 Rule-Based Classification
 Model Evaluation and Selection
 Techniques to Improve Classification Accuracy:
Ensemble Methods
 Summary
7
 Decision Tree Induction
 During the late 1970s and early 1980s, J. Ross Quinlan,
developed a decision tree algorithm known as ID3
(Iterative Dichotomiser).
 Quinlan later presented C4.5 (a successor of ID3),
 a group of statisticians (L. Breiman, J. Friedman, R.
Olshen, and C. Stone) published the book Classification
and Regression Trees (CART), which described the
generation of binary decision trees. ID3 and CART
were invented independently of one another at
around the same time, yet follow a similar approach
for learning decision trees from training tuples.

8
 Decision Tree Induction: An Example
age income student credit_rating buys_computer
<=30 high no fair no
 Training data set: Buys_computer <=30 high no excellent no
 The data set follows an example of 31…40 high no fair yes
>40 medium no fair yes
Quinlan’s ID3 (Playing Tennis) >40 low yes fair yes
 Resulting tree: >40 low yes excellent no
31…40 low yes excellent yes
age? <=30 medium no fair no
<=30 low yes fair yes
>40 medium yes fair yes
<=30 medium yes excellent yes
<=30 overcast
31..40 >40 31…40 medium no excellent yes
31…40 high yes fair yes
>40 medium no excellent no

student? yes credit rating?

no yes excellent fair

no yes no yes
9
 Algorithm for Decision Tree Induction
 Basic algorithm (a greedy algorithm)
 Tree is constructed in a top-down recursive divide-and-

conquer manner
 At start, all the training examples are at the root

 Attributes are categorical (if continuous-valued, they are

discretized in advance)
 Examples are partitioned recursively based on selected

attributes
 Test attributes are selected on the basis of a heuristic or

statistical measure (e.g., information gain)

 Conditions for stopping partitioning
 All samples for a given node belong to the same class

 There are no remaining attributes for further partitioning –

majority voting is employed for classifying the leaf

 There are no samples left
10
 Brief Review of Entropy
  

m=2

11
 Attribute Selection Measure:
Information Gain (ID3/C4.5)
 Select the attribute with the highest information gain
 Let pi be the probability that an arbitrary tuple in D belongs to
class Ci, estimated by |Ci, D|/|D|
 Expected information (entropy) needed to classify
m a tuple in D:
Info ( D )    pi log 2 ( pi )
i 1
 Information needed (after using A to split D into v partitions) to
v | D |
classify D:
Info A ( D )  
j
 Info ( D j )
j 1 | D |
 Information gained by branching on attribute A
Gain(A)  Info(D)  Info A(D)
12
 Attribute Selection: Information Gain
g Class P: buys_computer = “yes” 5 4
Info age ( D )  I ( 2 ,3 )  I ( 4 ,0 )
g Class N: buys_computer = “no” 14 14
9 9 5 5 5
Info ( D )  I (9,5)   log 2 ( )  log 2 ( )  0 .940  I (3, 2 )  0 .694
14 14 14 14 14
age pi ni I(p i, n i) 5
<=30 2 3 0.971 I ( 2,3)means “age <=30” has 5 out of
14
31…40 4 0 0 14 samples, with 2 yes’es and 3
>40 3 2 0.971 no’s. Hence
age
<=30
income student credit_rating
high no fair
buys_computer
no
Gain( age)  Info( D )  Infoage ( D )  0.246
<=30 high no excellent no
31…40
>40
high
medium
no
no
fair
fair
yes
yes
Similarly,
>40 low yes fair yes
>40 low yes excellent no
31…40 low yes excellent yes Gain(income)  0.029
<=30 medium no fair no
<=30
>40
low
medium
yes
yes
fair
fair
yes
yes
Gain( student )  0.151
<=30
31…40
medium
medium
yes
no
excellent
excellent
yes
yes Gain(credit _ rating )  0.048
31…40 high yes fair yes
>40 medium no excellent no 13
 Computing Information-Gain for
Continuous-Valued Attributes
 Let attribute A be a continuous-valued attribute
 Must determine the best split point for A
 Sort the value A in increasing order
 Typically, the midpoint between each pair of adjacent values
is considered as a possible split point
 (ai+ai+1)/2 is the midpoint between the values of ai and ai+1
 The point with the minimum expected information
requirement for A is selected as the split-point for A
 Split:
 D1 is the set of tuples in D satisfying A ≤ split-point, and D2 is
the set of tuples in D satisfying A > split-point
14
 Gain Ratio for Attribute Selection (C4.5)
 Information gain measure is biased towards attributes with a
large number of values
 C4.5 (a successor of ID3) uses gain ratio to overcome the
problem (normalization to information gain)
v | Dj | | Dj |
SplitInfoA ( D)    log 2 ( )
j 1 |D| |D|
 GainRatio(A) = Gain(A)/SplitInfo(A)
 Ex.

 gain_ratio(income) = 0.029/1.557 = 0.019

 The attribute with the maximum gain ratio is selected as the
splitting attribute
15
 Gini Index (CART, IBM
IntelligentMiner)
 If a data set D contains examples from n classes, gini index,
gini(D) is defined as n 2
gini ( D )  1   p j
j 1
where pj is the relative frequency of class j in D
 If a data set D is split on A into two subsets D1 and D2, the gini
index gini(D) is defined as | D1 | |D2 |
gini A ( D )  gini ( D 1)  gini ( D 2 )
|D | |D |
 Reduction in Impurity:
gini( A)  gini(D)  giniA (D)
 The attribute provides the smallest ginisplit(D) (or the largest
reduction in impurity) is chosen to split the node (need to
enumerate all the possible splitting points for each attribute)
16
 Computation of Gini Index
 Ex. D has 9 tuples in buys_computer = “yes”
2
and
2
5 in “no”
9 5
gini( D)  1        0.459
 14   14 
 Suppose the attribute income partitions D into 10 in D1: {low,
10 4
medium} and 4 in D2 giniincome{low,medium} ( D)   Gini ( D1 )   Gini( D2 )
 14   14 

Gini{low,high} is 0.458; Gini{medium,high} is 0.450. Thus, split on the

{low,medium} (and {high}) since it has the lowest Gini index
 All attributes are assumed continuous-valued
 May need other tools, e.g., clustering, to get the possible split
values
 Can be modified for categorical attributes 17
 Comparing Attribute Selection Measures

 The three measures, in general, return good results but

 Information gain:
 biased towards multivalued attributes
 Gain ratio:
 tends to prefer unbalanced splits in which one partition is
much smaller than the others
 Gini index:
 biased to multivalued attributes
 has difficulty when # of classes is large
 tends to favor tests that result in equal-sized partitions
and purity in both partitions
18
 Other Attribute Selection Measures
 CHAID: a popular decision tree algorithm, measure based on χ2 test for
independence
 C-SEP: performs better than info. gain and gini index in certain cases
 G-statistic: has a close approximation to χ2 distribution
 MDL (Minimal Description Length) principle (i.e., the simplest solution is
preferred):
 The best tree as the one that requires the fewest # of bits to both (1)
encode the tree, and (2) encode the exceptions to the tree
 Multivariate splits (partition based on multiple variable combinations)
 CART: finds multivariate splits based on a linear comb. of attrs.
 Which attribute selection measure is the best?
 Most give good results, none is significantly superior than others
19
 Overfitting and Tree Pruning
 Overfitting: An induced tree may overfit the training data
 Too many branches, some may reflect anomalies due to

noise or outliers
 Poor accuracy for unseen samples

 Two approaches to avoid overfitting

 Prepruning: Halt tree construction early ̵ do not split a node

if this would result in the goodness measure falling below a

threshold
 Difficult to choose an appropriate threshold

 Postpruning: Remove branches from a “fully grown” tree—

get a sequence of progressively pruned trees

 Use a set of data different from the training data to

decide which is the “best pruned tree”

20
Enhancements to Basic Decision Tree Induction

 Allow for continuous-valued attributes

 Dynamically define new discrete-valued attributes that
partition the continuous attribute value into a discrete set of
intervals
 Handle missing attribute values
 Assign the most common value of the attribute
 Assign probability to each of the possible values
 Attribute construction
 Create new attributes based on existing ones that are
sparsely represented
 This reduces fragmentation, repetition, and replication

21
 Classification in Large Databases
 Classification—a classical problem extensively studied by
statisticians and machine learning researchers
 Scalability: Classifying data sets with millions of examples and
hundreds of attributes with reasonable speed
 Why is decision tree induction popular?
 relatively faster learning speed (than other classification

methods)
 convertible to simple and easy to understand classification

rules
 can use SQL queries for accessing databases

 comparable classification accuracy with other methods

 RainForest (VLDB’98 — Gehrke, Ramakrishnan & Ganti)

 Builds an AVC-list (attribute, value, class label)

22
 Scalability Framework for RainForest

 Separates the scalability aspects from the criteria that

determine the quality of the tree
 Builds an AVC-list: AVC (Attribute, Value, Class_label)
 AVC-set (of an attribute X )
 Projection of training dataset onto the attribute X and
class label where counts of individual class label are
aggregated
 AVC-group (of a node n )
 Set of AVC-sets of all predictor attributes at the node n

23
 Rainforest: Training Set and Its AVC Sets

Training Examples AVC-set on Age AVC-set on income

age income studentcredit_rating
buys_computerAge Buy_Computer income Buy_Computer

<=30 high no fair no yes no

yes no
<=30 high no excellent no
high 2 2
31…40 high no fair yes <=30 2 3
31..40 4 0 medium 4 2
>40 medium no fair yes
>40 low yes fair yes >40 3 2 low 3 1

>40 low yes excellent no

31…40 low yes excellent yes
AVC-set on
<=30 medium no fair no AVC-set on Student
credit_rating
<=30 low yes fair yes
student Buy_Computer
>40 medium yes fair yes Buy_Computer

<=30 medium yes excellent yes yes no Credit

rating yes no
31…40 medium no excellent yes yes 6 1
fair 6 2
31…40 high yes fair yes no 3 4
excellent 3 3
>40 medium no excellent no
24
 BOAT (Bootstrapped Optimistic
Algorithm for Tree Construction)
 Use a statistical technique called bootstrapping to create
several smaller samples (subsets), each fits in memory
 Each subset is used to create a tree, resulting in several
trees
 These trees are examined and used to construct a new
tree T’
 It turns out that T’ is very close to the tree that would
be generated using the whole data set together
 Adv: requires only two scans of DB, an incremental alg.

25
 Presentation of Classification Results

3/5/23 Data Mining: Concepts and Techniques 26

Visualization of a Decision Tree in SGI/MineSet 3.0

3/5/23 Data Mining: Concepts and Techniques 27

Interactive Visual Mining by Perception-
Based Classification (PBC)

Data Mining: Concepts and Techniques 28

 Classification: Basic Concepts
 Decision Tree Induction
 Bayes Classification Methods
 Rule-Based Classification
 Model Evaluation and Selection
 Techniques to Improve Classification Accuracy:
Ensemble Methods
 Summary
29
 Bayesian Classification: Why?
 A statistical classifier: performs probabilistic prediction, i.e.,
predicts class membership probabilities
 Foundation: Based on Bayes’ Theorem.
 Performance: A simple Bayesian classifier, naïve Bayesian
classifier, has comparable performance with decision tree and
selected neural network classifiers
 Incremental: Each training example can incrementally
increase/decrease the probability that a hypothesis is correct —
prior knowledge can be combined with observed data
 Standard: Even when Bayesian methods are computationally
intractable, they can provide a standard of optimal decision
making against which other methods can be measured
30
 Bayes’ Theorem: Basics
M
 Total probability Theorem: P(B)   P(B | A )P( A )
i i
i 1

 Bayes’ Theorem: P ( H | X )  P ( X | H ) P ( H )  P ( X | H )  P ( H ) / P ( X )
P (X )
 Let X be a data sample (“evidence”): class label is unknown
 Let H be a hypothesis that X belongs to class C
 Classification is to determine P(H|X), (i.e., posteriori probability): the
probability that the hypothesis holds given the observed data sample X
 P(H) (prior probability): the initial probability
 E.g., X will buy computer, regardless of age, income, …

 P(X): probability that sample data is observed

 P(X|H) (likelihood): the probability of observing the sample X, given that
the hypothesis holds
 E.g., Given that X will buy computer, the prob. that X is 31..40,

medium income
31
 Prediction Based on Bayes’ Theorem
 Given training data X, posteriori probability of a hypothesis H,
P(H|X), follows the Bayes’ theorem

P ( H | X )  P (X | H ) P ( H )  P (X | H ) P (H ) / P (X )
P (X )
 Informally, this can be viewed as
posteriori = likelihood x prior/evidence
 Predicts X belongs to Ci iff the probability P(Ci|X) is the highest
among all the P(Ck|X) for all the k classes
 Practical difficulty: It requires initial knowledge of many
probabilities, involving significant computational cost

32
Classification Is to Derive the Maximum Posteriori
 Let D be a training set of tuples and their associated class labels,
and each tuple is represented by an n-D attribute vector X = (x1,
x2, …, xn)
 Suppose there are m classes C1, C2, …, Cm.
 Classification is to derive the maximum posteriori, i.e., the
maximal P(Ci|X)
 This can be derived from Bayes’ theorem
P(X | C )P(C )
P(C | X)  i i
i P(X)

 Since P(X) is constant for all classes, only

P(C | X)  P(X | C )P(C )
i i i
needs to be maximized

33
 Naïve Bayes Classifier
 A simplified assumption: attributes are conditionally
independent (i.e., no dependence relation between attributes):
n
P(X | Ci)   P(x | Ci)  P(x | Ci)  P(x | Ci) ... P(x | Ci)
k 1
 This greatly reduces thek computation
1 cost: Only2counts the n
class distribution
 If Ak is categorical, P(xk|Ci) is the # of tuples in Ci having value xk
for Ak divided by |Ci, D| (# of tuples of Ci in D)
 If Ak is continous-valued, P(xk|Ci) is usually computed based on
Gaussian distribution with a mean μ and standard deviation σ
( x  )2
and P(xk|Ci) is 1 
g ( x,  ,  ) 
2
e 2

2 

P(X| Ci)  g(xk , Ci ,Ci )

34
 Naïve Bayes Classifier: Training Dataset
age income studentcredit_rating
buys_compu
<=30 high no fair no
Class: <=30 high no excellent no
C1:buys_computer = ‘yes’ 31…40 high no fair yes
C2:buys_computer = ‘no’ >40 medium no fair yes
>40 low yes fair yes
>40 low yes excellent no
Data to be classified:
31…40 low yes excellent yes
X = (age <=30,
<=30 medium no fair no
Income = medium, <=30 low yes fair yes
Student = yes >40 medium yes fair yes
Credit_rating = Fair) <=30 medium yes excellent yes
31…40 medium no excellent yes
31…40 high yes fair yes
>40 medium no excellent no
35
 Naïve Bayes Classifier: An Example age income studentcredit_rating
buys_comp
<=30 high no fair no
<=30 high no excellent no
31…40 high no fair yes
 P(Ci): P(buys_computer = “yes”) = 9/14 = 0.643 >40
>40
>40
medium
low
low
no fair
yes fair
yes excellent
yes
yes
no

P(buys_computer = “no”) = 5/14= 0.357 31…40

<=30
low
medium
yes excellent
no fair
yes
no
<=30 low yes fair yes
 Compute P(X|Ci) for each class >40
<=30
31…40
medium yes fair
medium yes excellent
medium no excellent
yes
yes
yes

P(age = “<=30” | buys_computer = “yes”) = 2/9 = 0.222 31…40

>40
high
medium
yes fair
no excellent
yes
no

P(age = “<= 30” | buys_computer = “no”) = 3/5 = 0.6

P(income = “medium” | buys_computer = “yes”) = 4/9 = 0.444
P(income = “medium” | buys_computer = “no”) = 2/5 = 0.4
P(student = “yes” | buys_computer = “yes) = 6/9 = 0.667
P(student = “yes” | buys_computer = “no”) = 1/5 = 0.2
P(credit_rating = “fair” | buys_computer = “yes”) = 6/9 = 0.667
P(credit_rating = “fair” | buys_computer = “no”) = 2/5 = 0.4
 X = (age <= 30 , income = medium, student = yes, credit_rating = fair)
P(X|Ci) : P(X|buys_computer = “yes”) = 0.222 x 0.444 x 0.667 x 0.667 = 0.044
P(X|buys_computer = “no”) = 0.6 x 0.4 x 0.2 x 0.4 = 0.019
P(X|Ci)*P(Ci) : P(X|buys_computer = “yes”) * P(buys_computer = “yes”) = 0.028
P(X|buys_computer = “no”) * P(buys_computer = “no”) = 0.007
Therefore, X belongs to class (“buys_computer = yes”) 36
 Avoiding the Zero-Probability Problem
 Naïve Bayesian prediction requires each conditional prob. be
non-zero. Otherwise, the predicted prob. will be zero
n
P ( X | C i)   P ( x k | C i)
k 1
 Ex. Suppose a dataset with 1000 tuples, income=low (0),
income= medium (990), and income = high (10)
 Use Laplacian correction (or Laplacian estimator)
 Adding 1 to each case

Prob(income = low) = 1/1003

Prob(income = medium) = 991/1003
Prob(income = high) = 11/1003
 The “corrected” prob. estimates are close to their

“uncorrected” counterparts
37
 Naïve Bayes Classifier: Comments
 Advantages
 Easy to implement

 Good results obtained in most of the cases

 Disadvantages
 Assumption: class conditional independence, therefore loss of

accuracy
 Practically, dependencies exist among variables

 E.g., hospitals: patients: Profile: age, family history, etc.

Symptoms: fever, cough etc., Disease: lung cancer,

diabetes, etc.
 Dependencies among these cannot be modeled by Naïve

Bayes Classifier
 How to deal with these dependencies? Bayesian Belief Networks
(Chapter 9)
38
 Classification: Basic Concepts
 Decision Tree Induction
 Bayes Classification Methods
 Rule-Based Classification
 Model Evaluation and Selection
 Techniques to Improve Classification Accuracy:
Ensemble Methods
 Summary
39
 Using IF-THEN Rules for Classification
 Represent the knowledge in the form of IF-THEN rules
R: IF age = youth AND student = yes THEN buys_computer = yes
 Rule antecedent/precondition vs. rule consequent

 Assessment of a rule: coverage and accuracy

 n
covers = # of tuples covered by R

 ncorrect = # of tuples correctly classified by R

coverage(R) = ncovers /|D| /* D: training data set */
accuracy(R) = ncorrect / ncovers
 If more than one rule are triggered, need conflict resolution
 Size ordering: assign the highest priority to the triggering rules that has

the “toughest” requirement (i.e., with the most attribute tests)

 Class-based ordering: decreasing order of prevalence or misclassification

cost per class

 Rule-based ordering (decision list): rules are organized into one long

priority list, according to some measure of rule quality or by experts

40
 Rule Extraction from a Decision Tree
 Rules are easier to understand than large
trees age?
 One rule is created for each path from the <=30 31..40 >40
root to a leaf
student? credit rating?
yes
 Each attribute-value pair along a path forms a
excellent fair
conjunction: the leaf holds the class no yes

no yes
prediction no yes

 Rules are mutually exclusive and exhaustive

 Example: Rule extraction from our buys_computer decision-tree
IF age = young AND student = no THEN buys_computer = no
IF age = young AND student = yes THEN buys_computer = yes
IF age = mid-age THEN buys_computer = yes
IF age = old AND credit_rating = excellent THEN buys_computer = no
IF age = old AND credit_rating = fair THEN buys_computer = yes
41
Rule Induction: Sequential Covering Method
 Sequential covering algorithm: Extracts rules directly from training
data
 Typical sequential covering algorithms: FOIL, AQ, CN2, RIPPER
 Rules are learned sequentially, each for a given class Ci will cover
many tuples of Ci but none (or few) of the tuples of other classes
 Steps:
 Rules are learned one at a time

 Each time a rule is learned, the tuples covered by the rules are

removed
 Repeat the process on the remaining tuples until termination

condition, e.g., when no more training examples or when the

quality of a rule returned is below a user-specified threshold
 Comp. w. decision-tree induction: learning a set of rules
simultaneously
42
 Sequential Covering Algorithm

while (enough target tuples left)

generate a rule
remove positive target tuples satisfying this rule

Examples covered
Examples covered by Rule 2
by Rule 1 Examples covered
by Rule 3

Positive
examples

43
 Rule Generation
 To generate a rule
while(true)
find the best predicate p
if foil-gain(p) > threshold then add p to current rule
else break

A3=1&&A1=2
A3=1&&A1=2
&&A8=5A3=1

Positive Negative
examples examples

44
 How to Learn-One-Rule?
 Start with the most general rule possible: condition = empty
 Adding new attributes by adopting a greedy depth-first strategy
 Picks the one that most improves the rule quality

 Rule-Quality measures: consider both coverage and accuracy

 Foil-gain (in FOIL & RIPPER): assesses info_gain by extending

condition pos' pos

FOIL_ Gain pos'(log2  log2 )
pos'neg' pos neg
 favors rules that have high accuracy and cover many positive tuples
 Rule pruning based on an independent set of test tuples
pos  neg
FOIL _ Prune( R) 
pos  neg
Pos/neg are # of positive/negative tuples covered by R.
If FOIL_Prune is higher for the pruned version of R, prune R
45
 Classification: Basic Concepts
 Decision Tree Induction
 Bayes Classification Methods
 Rule-Based Classification
 Model Evaluation and Selection
 Techniques to Improve Classification Accuracy:
Ensemble Methods
 Summary
46
 Model Evaluation and Selection
 Evaluation metrics: How can we measure accuracy? Other
metrics to consider?
 Use validation test set of class-labeled tuples instead of training
set when assessing accuracy
 Methods for estimating a classifier’s accuracy:
 Holdout method, random subsampling
 Cross-validation
 Bootstrap
 Comparing classifiers:
 Confidence intervals
 Cost-benefit analysis and ROC Curves

47
 Classifier Evaluation Metrics: Confusion
Matrix
Confusion Matrix:
Actual class\Predicted class C1 ¬ C1
C1 True Positives (TP) False Negatives (FN)
¬ C1 False Positives (FP) True Negatives (TN)

Example of Confusion Matrix:

Actual class\Predicted buy_computer buy_computer Total
class = yes = no
buy_computer = yes 6954 46 7000
buy_computer = no 412 2588 3000
Total 7366 2634 10000

 Given m classes, an entry, CMi,j in a confusion matrix indicates

# of tuples in class i that were labeled by the classifier as class j
 May have extra rows/columns to provide totals
48
 Classifier Evaluation Metrics: Accuracy,
Error Rate, Sensitivity and Specificity
A\P C ¬C  Class Imbalance Problem:
C TP FN P
 One class may be rare, e.g.
¬C FP TN N
fraud, or HIV-positive
P’ N’ All
 Significant majority of the

 Classifier Accuracy, or negative class and minority of

recognition rate: percentage of the positive class
test set tuples that are correctly  Sensitivity: True Positive
classified recognition rate
 Sensitivity = TP/P
Accuracy = (TP + TN)/All
 Specificity: True Negative
 Error rate: 1 – accuracy, or
Error rate = (FP + FN)/All recognition rate
 Specificity = TN/N

49
 Classifier Evaluation Metrics:
Precision and Recall, and F-measures
 Precision: exactness – what % of tuples that the classifier
labeled as positive are actually positive

 Recall: completeness – what % of positive tuples did the

classifier label as positive?
 Perfect score is 1.0
 Inverse relationship between precision & recall
 F measure (F1 or F-score): harmonic mean of precision and
recall,

 Fß: weighted measure of precision and recall

 assigns ß times as much weight to recall as to precision

50
 Classifier Evaluation Metrics: Example

Actual Class\Predicted class cancer = yes cancer = no Total Recognition(%)

cancer = yes 90 210 300 30.00 (sensitivity
cancer = no 140 9560 9700 98.56 (specificity)
Total 230 9770 10000 96.40 (accuracy)

 Precision = 90/230 = 39.13% Recall = 90/300 = 30.00%

51
 Evaluating Classifier Accuracy:
Holdout & Cross-Validation Methods
 Holdout method
 Given data is randomly partitioned into two independent sets

 Training set (e.g., 2/3) for model construction

 Test set (e.g., 1/3) for accuracy estimation

 Random sampling: a variation of holdout

 Repeat holdout k times, accuracy = avg. of the accuracies

obtained
 Cross-validation (k-fold, where k = 10 is most popular)
 Randomly partition the data into k mutually exclusive subsets,

each approximately equal size


At i-th iteration, use Di as test set and others as training set
 Leave-one-out: k folds where k = # of tuples, for small sized

data
 *Stratified cross-validation*: folds are stratified so that class

dist. in each fold is approx. the same as that in the initial data
52
Evaluating Classifier Accuracy: Bootstrap
 Bootstrap
 Works well with small data sets
 Samples the given training tuples uniformly with replacement
 i.e., each time a tuple is selected, it is equally likely to be selected
again and re-added to the training set
 Several bootstrap methods, and a common one is .632 boostrap
 A data set with d tuples is sampled d times, with replacement, resulting in
a training set of d samples. The data tuples that did not make it into the
training set end up forming the test set. About 63.2% of the original data
end up in the bootstrap, and the remaining 36.8% form the test set (since
(1 – 1/d)d ≈ e-1 = 0.368)
 Repeat the sampling procedure k times, overall accuracy of the model:

53
 Estimating Confidence Intervals:
Classifier Models M1 vs. M2
 Suppose we have 2 classifiers, M1 and M2, which one is better?
 Use 10-fold cross-validation to obtain and
 These mean error rates are just estimates of error on the true
population of future data cases
 What if the difference between the 2 error rates is just
attributed to chance?
 Use a test of statistical significance
 Obtain confidence limits for our error estimates

54
 Estimating Confidence Intervals:
Null Hypothesis
 Perform 10-fold cross-validation
 Assume samples follow a t distribution with k–1 degrees of
freedom (here, k=10)
 Use t-test (or Student’s t-test)
 Null Hypothesis: M1 & M2 are the same
 If we can reject null hypothesis, then
 we conclude that the difference between M1 & M2 is
statistically significant
 Chose model with lower error rate

55
 Estimating Confidence Intervals: t-test

 If only 1 test set available: pairwise comparison

 For ith round of 10-fold cross-validation, the same cross
partitioning is used to obtain err(M1)i and err(M2)i
 Average over 10 rounds to get and
 t-test computes t-statistic with k-1 degrees of
freedom:
where

 If two test sets available: use non-paired t-test

where

where k1 & k2 are # of cross-validation samples used for M1 & M2, resp.
56
 Estimating Confidence Intervals:
Table for t-distribution

 Symmetric
 Significance level,
e.g., sig = 0.05 or
5% means M1 & M2
are significantly
different for 95% of
population
 Confidence limit, z
= sig/2

57
 Estimating Confidence Intervals:
Statistical Significance
 Are M1 & M2 significantly different?
 Compute t. Select significance level (e.g. sig = 5%)
 Consult table for t-distribution: Find t value corresponding to
k-1 degrees of freedom (here, 9)
 t-distribution is symmetric: typically upper % points of
distribution shown → look up value for confidence limit
z=sig/2 (here, 0.025)
 If t > z or t < -z, then t value lies in rejection region:
 Reject null hypothesis that mean error rates of M & M
1 2
are same
 Conclude: statistically significant difference between M
1
& M2
 Otherwise, conclude that any difference is chance
58
Model Selection: ROC Curves
 ROC (Receiver Operating
Characteristics) curves: for visual
comparison of classification models
 Originated from signal detection theory
 Shows the trade-off between the true
positive rate and the false positive rate
 The area under the ROC curve is a  Vertical axis
measure of the accuracy of the model represents the true
positive rate
 Rank the test tuples in decreasing  Horizontal axis rep.
order: the one that is most likely to the false positive rate
belong to the positive class appears at  The plot also shows a
the top of the list diagonal line
 The closer to the diagonal line (i.e., the  A model with perfect
closer the area is to 0.5), the less accuracy will have an
accurate is the model area of 1.0
59
 Issues Affecting Model Selection
 Accuracy
 classifier accuracy: predicting class label
 Speed
 time to construct the model (training time)
 time to use the model (classification/prediction time)
 Robustness: handling noise and missing values
 Scalability: efficiency in disk-resident databases
 Interpretability
 understanding and insight provided by the model
 Other measures, e.g., goodness of rules, such as decision tree
size or compactness of classification rules
60
 Classification: Basic Concepts
 Decision Tree Induction
 Bayes Classification Methods
 Rule-Based Classification
 Model Evaluation and Selection
 Techniques to Improve Classification Accuracy:
Ensemble Methods
 Summary
61
 Ensemble Methods
 Problem Statement:
 Mining complex knowledge from complex data such as
from enterprise data mining solutions, has been one
of the most challenging problems in knowledge
discovery process.
 In such situations, it is observed that not a single
learning method able to provide best informative
knowledge in all kinds of datasets.
 The word ‘ensemble’ literally mean group.
 Ensemble methods involve group of predictive models
to achieve a better accuracy and model stability.
 Ensemble methods are known to impart supreme
 Bias and Variance
 Bias means, ‘how much on an average are the predicted values
different from the actual value.’
 Variance means, ‘how different will the predictions of the model be
at the same point if different samples are taken from the same
population’.
 You build a small tree and you will get a model with low variance
and high bias. How do you manage to balance the trade off
between bias and variance ?
 Increasing in the complexity of model, reduction in prediction error
due to lower bias in the model.
 As the model gets more complex, it end up with over-fitting i.e.
model will start suffering from high variance.
 Successful model is one that maintains a balance between these
two types of errors. This is known as the trade-off management of
bias-variance errors.
  Ensemble learning is one way to execute this trade off analysis.
Ensemble Methods: Increasing the Accuracy

 Ensemble methods
 Use a combination of models to increase accuracy

 Combine a series of k learned models, M , M , …, M , with

1 2 k
the aim of creating an improved model M*
 Popular ensemble methods
 Bagging: averaging the prediction over a collection of

classifiers
 Boosting: weighted vote with a collection of classifiers

 Ensemble: combining a set of heterogeneous classifiers

64
 Bagging: Boostrap Aggregation
 Analogy: Diagnosis based on multiple doctors’ majority vote
 Training
 Given a set D of d tuples, at each iteration i, a training set D of d tuples is
i
sampled with replacement from D (i.e., bootstrap)
 A classifier model M is learned for each training set D
i i
 Classification: classify an unknown sample X
 Each classifier M returns its class prediction
i
 The bagged classifier M* counts the votes and assigns the class with the
most votes to X
 Prediction: can be applied to the prediction of continuous values by taking
the average value of each prediction for a given test tuple
 Accuracy
 Often significantly better than a single classifier derived from D

 For noise data: not considerably worse, more robust

 Proved improved accuracy in prediction

65
 Boosting
 Analogy: Consult several doctors, based on a combination of
weighted diagnoses—weight assigned based on the previous
diagnosis accuracy
 How boosting works?
 Weights are assigned to each training tuple
 A series of k classifiers is iteratively learned
 After a classifier Mi is learned, the weights are updated to
allow the subsequent classifier, Mi+1, to pay more attention to
the training tuples that were misclassified by Mi
 The final M* combines the votes of each individual classifier,
where the weight of each classifier's vote is a function of its
accuracy
 Boosting algorithm can be extended for numeric prediction
 Comparing with bagging: Boosting tends to have greater accuracy,
but it also risks overfitting the model to misclassified data 66
 Pseudocode of Bagging Algorithm

Note: Bagging works especially well for unstable algorithms , where a small
changes in the training data might result some significant changes in the output
classifiers.
• However, this algorithm would yield poor ensembles for stable algorithms.
• Stable algorithms include K-nearest neighbor, whereas decision tree, rule
learners and neural networks are considered to be unstable

67
 Adaboost (Freund and Schapire, 1997)
 Given a set of d class-labeled tuples, (X1, y1), …, (Xd, yd)
 Initially, all the weights of tuples are set the same (1/d)
 Generate k classifiers in k rounds. At round i,
 Tuples from D are sampled (with replacement) to form a training set Di
of the same size
 Each tuple’s chance of being selected is based on its weight
 A classification model Mi is derived from Di
 Its error rate is calculated using Di as a test set
 If a tuple is misclassified, its weight is increased, o.w. it is decreased
 Error rate: err(Xj) is the misclassification error of tuple Xj. Classifier Mi error
rate is the sum of the weights of the misclassified tuples:
d
error ( M i )  w
j
j  err ( X j )

 The weight of classifier Mi’s vote is 1  error ( M i )

log
error ( M i )
68
 Random Forest (Breiman 2001)
 Random Forest:
 Each classifier in the ensemble is a decision tree classifier and is

generated using a random selection of attributes at each node to

determine the split
 During classification, each tree votes and the most popular class is

returned
 Two Methods to construct Random Forest:
 Forest-RI (random input selection): Randomly select, at each node, F

attributes as candidates for the split at the node. The CART methodology
is used to grow the trees to maximum size
 Forest-RC (random linear combinations): Creates new attributes (or

features) that are a linear combination of the existing attributes (reduces

the correlation between individual classifiers)
 Comparable in accuracy to Adaboost, but more robust to errors and outliers
 Insensitive to the number of attributes selected for consideration at each
split, and faster than bagging or boosting
69
 Classification of Class-Imbalanced Data Sets
 Class-imbalance problem: Rare positive example but numerous
negative ones, e.g., medical diagnosis, fraud, oil-spill, fault, etc.
 Traditional methods assume a balanced distribution of classes
and equal error costs: not suitable for class-imbalanced data
 Typical methods for imbalance data in 2-class classification:
 Oversampling: re-sampling of data from positive class

 Under-sampling: randomly eliminate tuples from negative

class
 Threshold-moving: moves the decision threshold, t, so that

the rare class tuples are easier to classify, and hence, less
chance of costly false negative errors
 Ensemble techniques: Ensemble multiple classifiers

introduced above
 Still difficult for class imbalance problem on multiclass tasks
70
 Classification: Basic Concepts
 Decision Tree Induction
 Bayes Classification Methods
 Rule-Based Classification
 Model Evaluation and Selection
 Techniques to Improve Classification Accuracy:
Ensemble Methods
 Summary
71
 Summary (I)
 Classification is a form of data analysis that extracts models
describing important data classes.
 Effective and scalable methods have been developed for decision
tree induction, Naive Bayesian classification, rule-based
classification, and many other classification methods.
 Evaluation metrics include: accuracy, sensitivity, specificity,
precision, recall, F measure, and Fß measure.
 Stratified k-fold cross-validation is recommended for accuracy
estimation. Bagging and boosting can be used to increase overall
accuracy by learning and combining a series of individual models.

72
 Summary (II)

 Significance tests and ROC curves are useful for model selection.
 There have been numerous comparisons of the different
classification methods; the matter remains a research topic
 No single method has been found to be superior over all others
for all data sets
 Issues such as accuracy, training time, robustness, scalability,
and interpretability must be considered and can involve trade-
offs, further complicating the quest for an overall superior
method

73