Data Mining Classification: Alternative Techniques

Data Mining
Classification: Alternative Techniques
Lecture Notes for Chapter 4
Rule-Based
Introduction to Data Mining , 2nd Edition

by
Tan, Steinbach, Karpatne, Kumar
Rule-Based Classifier
● Classify records by using a collection of

“if…then…” rules
● Rule: (Condition) → y
– where
u Condition is a conjunctions o f a ttributes
u y is the class label
– LHS: rule antecedent or condition
– RHS: rule consequent
– Examples of classification rules:
u (Blood Type=Warm) ∧ (Lay E ggs=Yes) → Birds
u (Taxable Income < 5 0K) ∧ (Refund=Yes) → Evade=No
02/14/2018 Introduction to Data Mining, 2nd Edition 2

Rule-based Classifier (Example)
Name Blood Type Give Birth Can Fly Live in Water Class
human warm yes no no mammals
python cold no no no reptiles
salmon cold no no yes fishes
whale warm yes no yes mammals
frog cold no no sometimes amphibians
komodo cold no no no reptiles
bat warm yes yes no mammals
pigeon warm no yes no birds
cat warm yes no no mammals
leopard shark cold yes no yes fishes
turtle cold no no sometimes reptiles
penguin warm no no sometimes birds
porcupine warm yes no no mammals
eel cold no no yes fishes
salamander cold no no sometimes amphibians
gila monster cold no no no reptiles
platypus warm no no no mammals
owl warm no yes no birds
dolphin warm yes no yes mammals
eagle warm no yes no birds
R1: (Give Birth = no) ∧ (Can Fly = yes) → Birds

R2: (Give Birth = no) ∧ (Live in Water = yes) → Fishes
R3: (Give Birth = yes) ∧ (Blood Type = warm) → Mammals
R4: (Give Birth = no) ∧ (Can Fly = no) → Reptiles
R5: (Live in Water = sometimes) → Amphibians
Application of Rule-Based Classifier
● A rule r covers an instance x if the attributes of

the instance satisfy the condition of the rule
R3: (Give Birth = y es) ∧ (Blood Type = warm) → Mammals
hawk warm no yes no ?
grizzly bear warm yes no no ?
The rule R1 covers a hawk => Bird

The rule R3 covers t he grizzly bear => Mammal

Rule Coverage and Accuracy
Tid Refund Marital Taxable
● Coverage of a rule: Status Income Class
– Fraction of records 1 Yes Single 125K No
that satisfy the 2 No Married 100K No

3 No Single 70K No
antecedent of a rule 4 Yes Married 120K No
● Accuracy of a rule: 5 No Divorced 95K Yes

6 No Married 60K No
– Fraction of records 7 Yes Divorced 220K No
that satisfy the 8 No Single 85K Yes
antecedent that 9 No Married 75K No
also satisfy the 10 No Single 90K Yes

10
consequent of a (Status=Single) → No

rule Coverage = 4 0%, Accuracy = 5 0%
How does Rule-based Classifier Work?

lemur warm yes no no ?
turtle cold no no sometimes ?
dogfish shark cold yes no yes ?
A lemur triggers rule R3, so it is c lassified as a mammal
A t urtle t riggers both R4 and R5
A dogfish s hark t riggers none of t he rules

Characteristics of Rule Sets: Strategy 1
● Mutually exclusive rules

– Classifier contains mutually exclusive rules if
the rules are independent of each other
– Every record is covered by at most one rule
● Exhaustive r ules
– Classifier has exhaustive coverage if it
accounts for every possible combination of
attribute values
– Each record is covered by at least one rule
Characteristics of Rule Sets: Strategy 2
● Rules are not m utually exclusive

– A record may trigger more than one rule
– Solution?
u Ordered rule set
u Unordered rule set – use voting schemes
● Rules are not exhaustive

– A record may not trigger any rules
– Solution?
u Use a default class

Ordered Rule Set
● Rules are r ank ordered according to their priority

– An o rdered rule set is known a s a d ecision list
● When a test record is presented to the classifier
– It is assigned to the class label o f the h ighest ranked rule it h as
triggered
– If n one o f the rules fired, it is assigned to the d efault class

turtle cold no no sometimes ?
Rule Ordering Schemes
● Rule-based ordering
– Individual rules a re ranked b ased o n their q uality
● Class-based ordering
– Rules that b elong to the same class a ppear together
Rule-based Ordering Class-based Ordering

(Refund=Yes) ==> No (Refund=Yes) ==> No
(Refund=No, Marital Status={Single,Divorced}, (Refund=No, Marital Status={Single,Divorced},

Taxable Income<80K) ==> No Taxable Income<80K) ==> No
(Refund=No, Marital Status={Single,Divorced}, (Refund=No, Marital Status={Married}) ==> No

Taxable Income>80K) ==> Yes
(Refund=No, Marital Status={Single,Divorced},
(Refund=No, Marital Status={Married}) ==> No Taxable Income>80K) ==> Yes

Building Classification Rules
● Direct Method:
u Extract rules directly from data
u Examples: RIPPER, CN2, Holte’s 1R
● Indirect M ethod:
u Extract rules from other classification models (e.g.
decision trees, neural networks, etc).
u Examples: C4.5rules
Direct Method: Sequential Covering
1. Start from an empty rule

2. Grow a rule using the Learn-One-Rule function
3. Remove training records covered by the rule
4. Repeat Step (2) and (3) until stopping criterion
is met

Example of Sequential Covering
(i) Original Data (ii) Step 1
Example of Sequential Covering…
R1 R1
R2
(iii) Step 2 (iv) Step 3

Instance Elimination
● Why do we need to

eliminate instances?
R3 R2
– Otherwise, the n ext rule is
identical to p revious rule R1 + + + +
+
+ ++ +
● Why do we remove class = +
+
+++
+ +
+
+
+
+ +
positive instances? + + + +
+ + +
+ +
– Ensure that the n ext rule is -
- -
-
- - -
-
different - -
class = - - - -
● Why do we remove - -
-
negative instances? -
- -
-
– Prevent u nderestimating -
accuracy o f rule
– Compare rules R2 and R3
in the d iagram
Rule Growing
● Two common strategies
Yes: 3
{ } No: 4
Refund=No, Refund=No,
Status=Single, Status=Single,
Income=85K Income=90K
(Class=Yes) (Class=Yes)
Refund=
No
Status =
Single
Status =
Divorced
Status =
Married
... Income
> 80K
Refund=No,
Status = Single
Yes: 3 Yes: 2 Yes: 1 Yes: 0 Yes: 3 (Class = Yes)
No: 4 No: 1 No: 0 No: 3 No: 1
(a) General-to-specific (b) Specific-to-general

Rule Evaluation
FOIL: First Order Inductive
● Foil’s Information Gain Learner – an early rule-
based learning algorithm
– R0: {} => class (initial rule)

– R1: {A} => class (rule after a dding conjunct)
– Gain(R0, R1) = t [ log (p1/(p1+n1)) – log (p0/(p0 + n 0)) ]
– where t: n umber o f p ositive instances covered b y

both R0 a nd R1
p0: n umber o f p ositive instances covered b y R0
n0: n umber o f n egative instances covered b y R0
p1: n umber o f p ositive instances covered b y R1
n1: n umber o f n egative instances covered b y R1
Direct Method: RIPPER
● For 2-class problem, choose one of the classes as

positive class, and the other as negative class
– Learn rules for positive class
– Negative class will be default class
● For multi-class problem
– Order the classes according to increasing class
prevalence (fraction of instances that belong to a
particular class)
– Learn the rule set for smallest class first, treat the rest
as negative class
– Repeat with next smallest class as positive class

● Growing a rule:
– Start from empty rule
– Add conjuncts as long as they improve FOIL’s
information gain
– Stop when rule no longer covers negative examples
– Prune the rule immediately using incremental reduced
error pruning
– Measure for pruning: v = (p-n)/(p+n)
u p: n umber o f p ositive e xamples covered b y the rule in
the validation set
u n: n umber o f n egative e xamples covered b y the rule in
the validation set
– Pruning method: delete any final sequence of
conditions that maximizes v
● Building a Rule Set:

– Use sequential covering algorithm
u Finds the best rule that covers the current set of
positive examples
u Eliminate both positive and negative examples
covered by the rule
– Each time a rule is added to the rule set,
compute the new description length
u Stop adding new rules when the new description
length is d bits longer than the smallest description
length obtained so far

● Optimize the rule set:

– For each rule r in the rule set R
u Consider 2 alternative rules:
– Replacement rule (r*): grow new rule from scratch
– Revised rule(r′): add conjuncts to extend the rule r
u Compare the rule set for r against the rule set for r*
and r′
u Choose rule set that minimizes MDL principle
– Repeat rule generation and rule optimization

for the remaining positive examples
Indirect Methods
P
No Yes
Q R Rule Set
No Yes No Yes r1: (P=No,Q=No) ==> -

r2: (P=No,Q=Yes) ==> +
- + + Q r3: (P=Yes,R=No) ==> +
r4: (P=Yes,R=Yes,Q=No) ==> -
No Yes
r5: (P=Yes,R=Yes,Q=Yes) ==> +
- +

Indirect Method: C4.5rules
● Extract r ules from an unpruned decision tree

● For each r ule, r : A → y,
– consider an alternative rule r′: A′ → y where A′
is obtained by removing one of the conjuncts
in A
– Compare the pessimistic error rate for r
against all r’s
– Prune if one of the alternative rules has lower
pessimistic error rate
– Repeat until we can no longer improve
generalization error
Indirect Method: C4.5rules
● Instead of ordering the r ules, order subsets of

rules (class ordering)
– Each subset is a collection of rules with the
same rule consequent (class)
– Compute description length of each subset
u Description length = L(error) + g L(model)
u g is a parameter that takes into a ccount the
presence of redundant attributes in a rule set
(default value = 0.5)

Example
Name Give Birth Lay Eggs Can Fly Live in Water Have Legs Class
human yes no no no yes mammals
python no yes no no no reptiles
salmon no yes no yes no fishes
whale yes no no yes no mammals
frog no yes no sometimes yes amphibians
komodo no yes no no yes reptiles
bat yes no yes no yes mammals
pigeon no yes yes no yes birds
cat yes no no no yes mammals
leopard shark yes no no yes no fishes
turtle no yes no sometimes yes reptiles
penguin no yes no sometimes yes birds
porcupine yes no no no yes mammals
eel no yes no yes no fishes
salamander no yes no sometimes yes amphibians
gila monster no yes no no yes reptiles
platypus no yes no no yes mammals
owl no yes yes no yes birds
dolphin yes no no yes no mammals
eagle no yes yes no yes birds
C4.5 versus C4.5rules versus RIPPER
Give C4.5rules:
Birth? (Give Birth=No, Can Fly=Yes) → Birds
(Give Birth=No, Live in Water=Yes) → Fishes
Yes No
(Give Birth=Yes) → Mammals
(Give Birth=No, Can Fly=No, Live in Water=No) → Reptiles
Mammals Live In ( ) → Amphibians
Water?
Yes No RIPPER:
(Live in Water=Yes) → Fishes
Sometimes (Have Legs=No) → Reptiles
(Give Birth=No, Can Fly=No, Live In Water=No)
Fishes Amphibians Can → Reptiles
Fly?
(Can Fly=Yes,Give Birth=No) → Birds
Yes No () → Mammals
Birds Reptiles

C4.5 versus C4.5rules versus RIPPER
C4.5 a nd C4.5rules:
PREDICTED CLASS
Amphibians Fishes Reptiles Birds Mammals
ACTUAL Amphibians 2 0 0 0 0
CLASS Fishes 0 2 0 0 1
Reptiles 1 0 3 0 0
Birds 1 0 0 3 0
Mammals 0 0 1 0 6
RIPPER:
PREDICTED CLASS
Amphibians Fishes Reptiles Birds Mammals
ACTUAL Amphibians 0 0 0 0 2
CLASS Fishes 0 3 0 0 0
Reptiles 0 0 3 0 1
Birds 0 0 1 2 1
Mammals 0 2 1 0 4
Advantages of Rule-Based Classifiers
● Has characteristics quite similar to decision trees

– As highly expressive as decision trees
– Easy to interpret
– Performance comparable to decision trees
– Can handle redundant attributes
● Better suited for handling imbalanced classes
● Harder to handle m issing values in the test set

Data Mining Classification: Alternative Techniques

Uploaded by

Copyright:

Available Formats

Data Mining Classification: Alternative Techniques

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Data Mining Classification: Alternative Techniques

Uploaded by

Copyright:

Available Formats

Data Mining

Classification: Alternative Techniques

Lecture Notes for Chapter 4

Introduction to Data Mining , 2nd Edition

● Classify records by using a collection of

02/14/2018 Introduction to Data Mining, 2nd Edition 2

R1: (Give Birth = no) ∧ (Can Fly = yes) → Birds

Application of Rule-­Based Classifier

● A rule r covers an instance x if the attributes of

The rule R1 covers a hawk => Bird

02/14/2018 Introduction to Data Mining, 2nd Edition 4

– Fraction of records 1 Yes Single 125K No

that satisfy the 2 No Married 100K No

● Accuracy of a rule: 5 No Divorced 95K Yes

antecedent that 9 No Married 75K No

also satisfy the 10 No Single 90K Yes

consequent of a (Status=Single) → No

02/14/2018 Introduction to Data Mining, 2nd Edition 5

How does Rule-­based Classifier Work?

R1: (Give Birth = no) ∧ (Can Fly = yes) → Birds

02/14/2018 Introduction to Data Mining, 2nd Edition 6

● Mutually exclusive rules

Characteristics of Rule Sets: Strategy 2

● Rules are not m utually exclusive

● Rules are not exhaustive

02/14/2018 Introduction to Data Mining, 2nd Edition 8

● Rules are r ank ordered according to their priority

R1: (Give Birth = no) ∧ (Can Fly = yes) → Birds

Rule Ordering Schemes

Rule-­based Ordering Class-­based Ordering

(Refund=No, Marital Status={Single,Divorced}, (Refund=No, Marital Status={Single,Divorced},

(Refund=No, Marital Status={Single,Divorced}, (Refund=No, Marital Status={Married}) ==> No

02/14/2018 Introduction to Data Mining, 2nd Edition 10

02/14/2018 Introduction to Data Mining, 2nd Edition 11

Direct Method: Sequential Covering

1. Start from an empty rule

02/14/2018 Introduction to Data Mining, 2nd Edition 12

(i) Original Data (ii) Step 1

02/14/2018 Introduction to Data Mining, 2nd Edition 13

Example of Sequential Covering…

(iii) Step 2 (iv) Step 3

02/14/2018 Introduction to Data Mining, 2nd Edition 14

● Why do we need to

02/14/2018 Introduction to Data Mining, 2nd Edition 15

● Two common strategies

(a) General-­to-­specific (b) Specific-­to-­general

02/14/2018 Introduction to Data Mining, 2nd Edition 16

– R0: {} => class (initial rule)

– Gain(R0, R1) = t [ log (p1/(p1+n1)) – log (p0/(p0 + n 0)) ]

– where t: n umber o f p ositive instances covered b y

02/14/2018 Introduction to Data Mining, 2nd Edition 17

Direct Method: RIPPER

● For 2-­class problem, choose one of the classes as

02/14/2018 Introduction to Data Mining, 2nd Edition 18

Direct Method: RIPPER

● Building a Rule Set:

02/14/2018 Introduction to Data Mining, 2nd Edition 20

● Optimize the rule set:

– Repeat rule generation and rule optimization

Application of Rule-Based Classifier

How does Rule-based Classifier Work?

Rule-based Ordering Class-based Ordering

(a) General-to-specific (b) Specific-to-general

● For 2-class problem, choose one of the classes as

No Yes No Yes r1: (P=No,Q=No) ==> -

Advantages of Rule-Based Classifiers