Machine Learning 2D5362

Machine Learning 2D5362

Lecture 3:
Decision Tree Learning

Homework: Concept Learner

S-boundary

G-boundary

Queries
guaranteed
to reduce the
version space
size

1

Homework: Concept Learner
Ideal training examples to teach the target concept

Outline

? Decision tree representation
? ID3 learning algorithm

? Entropy, information gain

? Overfitting

2

Decision Tree for PlayTennis
Outlook

Sunny Overcast Rain

Humidity Yes Wind

High Normal Strong Weak

No Yes No Yes

Outlook

Sunny Overcast Rain

Humidity Each internal node tests an attribute

High Normal Each branch corresponds to an
attribute value node
No Yes Each leaf node assigns a classification

3

Outlook Temperature Humidity Wind PlayTennis
Sunny Hot High Weak ?
No
Outlook

Sunny Overcast Rain

Humidity Yes Wind


No Yes No Yes

Decision Tree for Conjunction
Outlook=Sunny ? Wind=Weak

Outlook

Sunny Overcast Rain

Wind No No

Strong Weak

No Yes

4

Decision Tree for Disjunction
Outlook=Sunny ? Wind=Weak

Outlook

Sunny Overcast Rain

Yes Wind Wind

Strong Weak Strong Weak

No Yes No Yes

Decision Tree for XOR
Outlook=Sunny XOR Wind=Weak

Outlook

Sunny Overcast Rain

Wind Wind Wind

Strong Weak Strong Weak Strong Weak

Yes No No Yes No Yes

5

Decision Tree
• decision trees represent disjunctions of conjunctions
Outlook

Sunny Overcast Rain

Humidity Yes Wind

No Yes No Yes

(Outlook=Sunny ? Humidity=Normal)
? (Outlook=Overcast)
? (Outlook=Rain ? Wind=Weak)

When to consider Decision
Trees

? Instances describable by attribute-value pairs
? Target function is discrete valued
? Disjunctive hypothesis may be required
? Possibly noisy training data
? Missing attribute values
? Examples:
? Medical diagnosis
? Credit risk analysis
? Object classification for robot manipulator (Tan 1993)

6

Top-Down Induction of
Decision Trees ID3

1. A ? the “best” decision attribute for next node
2. Assign A as decision attribute for node
3. For each value of A create new descendant
4. Sort training examples to leaf node according to
the attribute value of the branch
5. If all training examples are perfectly classified
(same value of target attribute) stop, else
iterate over new leaf nodes.

Which Attribute is ”best”?

[29+,35-] A1=? A2=? [29+,35-]

True False True False

[21+, 5-] [8+, 30-] [18+, 33-] [11+, 2-]

7

Entropy

? S is a sample of training examples
? p+ is the proportion of positive examples
? p- is the proportion of negative examples
? Entropy measures the impurity of S
Entropy(S) = -p+ log 2 p+ - p- log 2 p-

Entropy
? Entropy(S)= expected number of bits needed to
encode class (+ or -) of randomly drawn
members of S (under the optimal, shortest
length-code)
Why?
? Information theory optimal length code assign

–log 2 p bits to messages having probability p.
? So the expected number of bits to encode

(+ or -) of random member of S:
-p+ log 2 p+ - p- log 2 p-

8

Information Gain
? Gain(S,A): expected reduction in entropy due
to sorting S on attribute A
Gain(S,A)=Entropy(S) - ? v? values(A) |Sv|/|S| Entropy(Sv)
Entropy([29+,35-]) = -29/64 log2 29/64 – 35/64 log2 35/64
= 0.99
[29+,35-] A1=? A2=? [29+,35-]


[21+, 5-] [8+, 30-] [18+, 33-] [11+, 2-]

Information Gain
Entropy([21+,5-]) = 0.71 Entropy([18+,33-]) = 0.94
Entropy([8+,30-]) = 0.74 Entropy([8+,30-]) = 0.62
Gain(S,A 1)=Entropy(S) Gain(S,A 2)=Entropy(S)
-26/64*Entropy([21+,5-]) -51/64*Entropy([18+,33-])
-38/64*Entropy([8+,30-]) -13/64*Entropy([11+,2-])
=0.27 =0.12

[29+,35-] A1=? A2=? [29+,35-]


[21+, 5-] [8+, 30-] [18+, 33-] [11+, 2-]

9

Training Examples
Day Outlook Temp. Humidity Wind Play Tennis
D1 Sunny Hot High Weak No
D2 Sunny Hot High Strong No
D3 Overcast Hot High Weak Yes
D4 Rain Mild High Weak Yes
D5 Rain Cool Normal Weak Yes
D6 Rain Cool Normal Strong No
D7 Overcast Cool Normal Weak Yes
D8 Sunny Mild High Weak No
D9 Sunny Cold Normal Weak Yes
D10 Rain Mild Normal Strong Yes
D11 Sunny Mild Normal Strong Yes
D12 Overcast Mild High Strong Yes
D13 Overcast Hot Normal Weak Yes
D14 Rain Mild High Strong No

Selecting the Next Attribute
S=[9+,5-] S=[9+,5-]
E=0.940 E=0.940
Humidity Wind

High Normal Weak Strong

[3+, 4-] [6+, 1-] [6+, 2-] [3+, 3-]
E=0.985 E=0.592 E=0.811 E=1.0
Gain(S,Humidity) Gain(S,Wind)
=0.940-(7/14)*0.985 =0.940-(8/14)*0.811
– (7/14)*0.592 – (6/14)*1.0
=0.151 =0.048

10

Selecting the Next Attribute
S=[9+,5-]
E=0.940
Outlook

Over
Sunny Rain
cast

[2+, 3-] [4+, 0] [3+, 2-]
E=0.971 E=0.0 E=0.971
Gain(S,Outlook)
=0.940-(5/14)*0.971
-(4/14)*0.0 – (5/14)*0.0971
=0.247

ID3 Algorithm
[D1,D2,…,D14] Outlook
[9+,5-]

Sunny Overcast Rain

Ssunny=[D1,D2,D8,D9,D11] [D3,D7,D12,D13] [D4,D5,D6,D10,D14]
[2+,3-] [4+,0-] [3+,2-]
? Yes ?
Gain(Ssunny , Humidity)=0.970-(3/5)0.0 – 2/5(0.0) = 0.970
Gain(Ssunny , Temp.)=0.970-(2/5)0.0 –2/5(1.0)-(1/5)0.0 = 0.570
Gain(Ssunny , Wind)=0.970= -(2/5)1.0 – 3/5(0.918) = 0.019

11

ID3 Algorithm
Outlook

Sunny Overcast Rain

Humidity Yes Wind
[D3,D7,D12,D13]


No Yes No Yes

[D1,D2] [D8,D9,D11] [D6,D14] [D4,D5,D10]

Hypothesis Space Search ID3

+ - +
A2
A1
+ - + + - -
+ - + - - +
A2 A2

- + - + -
A3 A4
+ - - +

12

Hypothesis Space Search ID3
? Hypothesis space is complete!
? Target function surely in there…

? Outputs a single hypothesis
? No backtracking on selected attributes (greedy search)
? Local minimal (suboptimal splits)

? Statistically-based search choices
? Robust to noisy data

? Inductive bias (search bias)
? Prefer shorter trees over longer ones

? Place high information gain attributes close to the root

Inductive Bias in ID3

? H is the power set of instances X
? Unbiased ?

? Preference for short trees, and for those with high
information gain attributes near the root
? Bias is a preference for some hypotheses, rather than
a restriction of the hypothesis space H
? Occam’s razor: prefer the shortest (simplest)
hypothesis that fits the data

13

Occam’s Razor
Why prefer short hypotheses?
Argument in favor:
? Fewer short hypotheses than long hypotheses
? A short hypothesis that fits the data is unlikely to be a
coincidence
? A long hypothesis that fits the data might be a coincidence
Argument opposed:
? There are many ways to define small sets of hypotheses
? E.g. All trees with a prime number of nodes that use attributes
beginning with ”Z”
? What is so special about small sets based on size of hypothesis

Occam’s Razor
Definition A: time
green
blue
Hypothesis A:
Objects do not instantaneously change their color.
Definition B:
grue
bleen

Hypothesis B: 1.1.2000 0:00
On 1.1.2000 objects that were grue turned instantaneously
bleen and objects that were bleen turned instantaneously grue.

14

Overfitting

Consider error of hypothesis h over
? Training data: errortrain(h)

? Entire distribution D of data: errorD(h)

Hypothesis h? H overfits training data if there is
an alternative hypothesis h’? H such that
errortrain(h) < errortrain(h’)
and
errorD(h) > errorD(h’)

Overfitting in Decision Tree
Learning

15

Avoid Overfitting
How can we avoid overfitting?
? Stop growing when data split not statistically

significant
? Grow full tree then post-prune

? Minimum description length (MDL):

Minimize:
size(tree) + size(misclassifications(tree))

Reduced-Error Pruning

Split data into training and validation set
Do until further pruning is harmful:
1. Evaluate impact on validation set of pruning
each possible node (plus those below it)
2. Greedily remove the one that most
improves the validation set accuracy

Produces smallest version of most accurate
subtree

16

Effect of Reduced Error
Pruning

Rule-Post Pruning

1. Convert tree to equivalent set of rules
2. Prune each rule independently of each
other
3. Sort final rules into a desired sequence to
use

Method used in C4.5

17

Converting a Tree to Rules
Outlook

Sunny Overcast Rain

Humidity Yes Wind

No Yes No Yes
R1: If (Outlook=Sunny) ? (Humidity=High) Then PlayTennis=No
R2: If (Outlook=Sunny) ? (Humidity=Normal) Then PlayTennis=Yes
R3: If (Outlook=Overcast) Then PlayTennis=Yes
R4: If (Outlook=Rain) ? (Wind=Strong) Then PlayTennis=No
R5: If (Outlook=Rain) ? (Wind=Weak) Then PlayTennis=Yes

Continuous Valued Attributes
Create a discrete attribute to test continuous
? Temperature = 24.50C

? (Temperature > 20.00C) = {true, false}

Where to set the threshold?

Temperatur 150C 180C 190C 220C 240C 270C
PlayTennis No No Yes Yes Yes No

(see paper by [Fayyad, Irani 1993]

18

Attributes with many Values

? Problem: if an attribute has many values, maximizing
InformationGain will select it.
? E.g.: Imagine using Date=12.7.1996 as attribute

perfectly splits the data into subsets of size 1
Use GainRatio instead of information gain as criteria:
GainRatio(S,A) = Gain(S,A) / SplitInformation(S,A)
SplitInformation(S,A) = -? i=1..c |Si|/|S| log2 |Si|/|S|
Where Si is the subset for which attribute A has the value v i

Attributes with Cost

Consider:
? Medical diagnosis : blood test costs 1000 SEK

? Robotics: width_from_one_feet has cost 23 secs.

How to learn a consistent tree with low expected
cost?
Replace Gain by :
Gain2(S,A)/Cost(A) [Tan, Schimmer 1990]
2Gain(S,A)-1/(Cost(A)+1)w w ? [0,1] [Nunez 1988]

19

Unknown Attribute Values

What is some examples missing values of A?
Use training example anyway sort through tree
? If node n tests A, assign most common value of A among

other examples sorted to node n.
? Assign most common value of A among other examples

with same target value
? Assign probability pi to each possible value vi of A

? Assign fraction pi of example to each descendant in tree

Classify new examples in the same fashion

Cross-Validation
? Estimate the accuracy of a hypothesis
induced by a supervised learning algorithm
? Predict the accuracy of a hypothesis over
future unseen instances
? Select the optimal hypothesis from a given
set of alternative hypotheses
? Pruning decision trees
? Model selection
? Feature selection
? Combining multiple classifiers (boosting)

20

Holdout Method
? Partition data set D = {(v1,y1),…,(vn,yn)} into
training Dt and validation set Dh=DDt

Training Dt Validation DDt

acch = 1/h ? (vi,yi)? Dh ?(I(Dt,vi),yi)

I(Dt,vi) : output of hypothesis induced by learner I
trained on data Dt for instance v i
? (i,j) = 1 if i=j and 0 otherwise

Problems:
• makes insufficient use of data
• training and validation set are correlated

Cross-Validation
? k-fold cross-validation splits the data set D into
k mutually exclusive subsets D1,D2,…,Dk

D1 D2 D3 D4

? Train and test the learning algorithm k times,
each time it is trained on DDi and tested on D i
D1 D2 D3 D4 D1 D2 D3 D4

D1 D2 D3 D4 D1 D2 D3 D4

acccv = 1/n ? (vi,yi)? D ?(I(DDi,vi),yi)

21

Cross-Validation
? Uses all the data for training and testing
? Complete k-fold cross-validation splits the
dataset of size m in all (m over m/k) possible
ways (choosing m/k instances out of m)
? Leave n-out cross-validation sets n instances
aside for testing and uses the remaining ones
for training (leave one-out is equivalent to n-
fold cross-validation)
? In stratified cross-validation, the folds are
stratified so that they contain approximately
the same proportion of labels as the original
data set

Bootstrap
? Samples n instances uniformly from the data set
with replacement
? Probability that any given instance is not chosen
after n samples is (1-1/n)n ? e-1 ? 0.632
? The bootstrap sample is used for training the
remaining instances are used for testing
? accboot = 1/b ? i=1b (0.632 ?0i + 0.368 accs)
where ?0i is the accuracy on the test data of the
i-th bootstrap sample, accs is the accuracy
estimate on the training set and b the number of
bootstrap samples

22

Wrapper Model

Input Feature subset Induction
features search algorithm

Feature subset
evaluation

Feature subset
evaluation

Wrapper Model
? Evaluate the accuracy of the inducer for a given subset of
features by means of n-fold cross-validation
? The training data is split into n folds, and the induction
algorithm is run n times. The accuracy results are averaged
to produce the estimated accuracy.
? Forward elimination:
Starts with the empty set of features and greedily adds the
feature that improves the estimated accuracy at most
? Backward elimination:
Starts with the set of all features and greedily removes
features and greedily removes the worst feature

23

Bagging
? For each trial t=1,2,…,T create a bootstrap sample of size N.
? Generate a classifier C from the bootstrap sample
t

? The final classifier C* takes class that receives the majority

votes among the Ct
yes
instance C*

yes no yes
C1 C2 … CT

train train train

Training set1 Training set2 … Training setT

Bagging

? Bagging requires ”instable” classifiers like for
example decision trees or neural networks

”The vital element is the instability of the
prediction method. If perturbing the learning
set can cause significant changes in the
predictor constructed, then bagging can
improve accuracy.” (Breiman 1996)

24

MLC++
? Machine Learning library in C++ developed
by Ron Kohavi et al
? Includes decision trees (ID3), decision
graphs, nearest-neighbors, probabilities
(Naive-Bayes), perceptron
? Provides interfaces for external (not part of
MLC++) inducers such as C4.5, OC1, PEBLS,
CN2
? Includes a set of tools (wrappers) for
producing learning curves, perform accuracy
estimation, generate statistics, bagging,
feature subset selection

MLC++
? MLC++ comes in two versions
? MLC++ utilities 2.0: compiled code for SUN
takes options from the command line or
environment variables
? MLC++ source code: you build the libraries and
use the classes and methods in your own program
? Documentation:
? Manual for the MLC++ Utilities : decent
? Tutorial for MLC++ libraries: minimal, you find
some documentation in the source code itself

25

MLC++
Preliminaries:
? Use gcc 2.7.2.3 not gcc 2.95
? module add gcc/2.7.2.3

? MLC++ install directories
? /afs/nada.kth.se/home/cvap/hoffmann/Teaching/ML/MLC
? setenv MLCDIR /afs/nada.kth.se/home/cvap/hoffmann/ Teaching/ML/MLC
? source $MLCDIR/setup.SUN.GNU
? setenv PATH $PATH:$MLCDIR/external/gnu:$MLCDIR/bin
? Test installation
? cp $MLCDIR/tutorial/dt1.c $MLCDIR/tutorial/Makefile .
? make dt1
? ./dt1

Data Files
? One dataset has three associated files:
? <dataset>.names : describes how to parse the data
? <dataset>.data : contains the training instances
? <dataset>.test : contains additional instances for estimating
the accuracy
? Monk1.names ( | starts a comment)
no, yes
Head shape : round, square, octagon
Body shape : round, square, octagon
Holding : sword, ballon, flag
Jacket color: red, yellow, green, blue
Has tie: yes, no
? Monk1.data
round, round, yes, sword, green, yes, yes
round, round, yes, sword, green, no, yes

26

Example ID3

#include <basics.h> // Must be included first in any program using MLC++
#include <ID3Inducer.h> // Include ID3Inducer class declarations
main()
{
ID3Inducer inducer(”my_first_ID3"); // "inducer" is an object of ID3Inducer.
// The argument is the "description."
inducer.set_unknown_edges(FALSE);// Avoid edges for "unknown" attributes.
inducer.read_data(”monk"); // Read data files tutorial.names and
// tutorial.data
inducer.train(); // Produce categorizer
inducer.display_struct(); // Display structure of categorizer

return 0; // Return success to shell
}

Example Test Inducer
#include <ID3Inducer.h>
#include <PerfEstDispatch.h>
ID3Inducer inducer(”my_first_ID3");
inducer.set_unknown_edges(FALSE);
inducer.read_data(”monk");
Inducer.train();
InstanceList train(”monk1”);
InstanceList test(train.get_schema(),”monk1”,”.names”,”.test”);
CatTestResult result(inducer.get_categorizer(), train, test);
result.display();
Classifying (% done): 10% 20% 30% 40% 50% 60% 70% 80% 90
Number of training instances: 124
Number of test instances: 432. Unseen: 308, seen 124.
Number correct: 350. Number incorrect: 82
Generalization error: 26.62%. Memorization error: 0.00%
Error: 18.98% +- 1.89% [15.56% - 22.95%]
Average Normalized Mean Squared Error: 18.98%
Average Normalized Mean Absolute Error: 18.98%

27

Example Bootstrap Estimation
#include <Bootstrap.h>
#include <PerfEstDispatch.h>

InstanceList train(”monk1”);
Bootstrap bootstrap(20);
bootstrap.init_rand_num_ gen(0);
bootstrap.estimate_performance(inducer,train);
bootstrap.display();

Bootstrapping 20 times: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
13.30% +- 1.20% (2.81% - 22.05%)

Example Learning Curve
#include <LearnCurve.h>
LearnCurve learncurve ;
learncurve.set_sizes_and_trials (7, 20, 1, 120, train);
learncurve.init_rand_num_ gen(0);
learncurve.learn_curve(inducer,train);

Train Size: 1; training 20 times.................... Mean Error: 50.41% +- 0.00%

28

Example Bagging
ID3Inducer inducer(”my_first_id3");
BaggingInd baggind("Baggind");
CatTestResult resultid3(inducer.get_categorizer(), train, "monk1",
".names", ".test");
Mcout << resultid3 << endl;
baggind.set_main_inducer(&inducer);
baggind.read_data("monk1");
baggind.train();
CatTestResult resultbagg(baggind.get_categorizer(), train,
"monk1",".names",".test");
Mcout << resultbagg << endl;
Number of test instances: 432. Unseen: 308, seen 124.
Error: 18.98% +- 1.89% [15.56% - 22.95%]
Error: 10.19% +- 1.46% [7.67% - 13.40%]

Homework
? Read chapter 3 on decision tree learning
? Read at least one out the following three
articles
? ”A study of cross-validation and bootstrap for
accuracy estimation and model selection” [Kohavi
1995]
? ”Irrelevant features and the subset selection
problem” [John, Kohavi, Pfleger]
? ”Multi-Interval Discretization of Continuous-Valued
Attributes for Classification Learning” [Fayyad,
Irani 1993]

29

Machine Learning 2D5362

More Related Content

Similar to Machine Learning 2D5362 (15)

More from butest (20)

Machine Learning 2D5362