Decision Trees TJ Machine Learning Club Classification vs. - - PowerPoint PPT Presentation

▶
decision trees
SMART_READER_LITE
LIVE PREVIEW

Decision Trees TJ Machine Learning Club Classification vs. - - PowerPoint PPT Presentation

Dec 28, 2023 233 likes •506 views

Decision Trees TJ Machine Learning Club Classification vs. Regression Classification Classifying photos of fruits Determining whether tumor is benign or malignant Regression Predicting COVID-19 cases given demographic data

slide-1
SLIDE 1

Decision Trees

TJ Machine Learning Club

slide-2
SLIDE 2

Classification vs. Regression

○ Classification ■ Classifying photos of fruits ■ Determining whether tumor is benign or malignant ○ Regression ■ Predicting COVID-19 cases given demographic data ■ Predicting house prices given house features

Source: https://medium.com/datasoc/whats-the-problem-1ff8b338094b

slide-3
SLIDE 3

Features vs. Labels

Features (like x): Characteristics of the input

  • In the picture, features are

whether or not patient smokes (smoke), consumes alcohol (alco), and performs physical activity (active) Label (like y): The prediction or classification of the input

  • Whether or not patient has

cardiovascular disease (cardio)

Features Labels

slide-4
SLIDE 4

Training and Testing Datasets

Training data has both features and labels Testing data only has the features

Need to predict cardio

slide-5
SLIDE 5

What is a Decision Tree?

  • A decision tree is just a series of questions
  • The key in creating a decision tree is asking the right questions
slide-6
SLIDE 6

Gini Impurity

  • Measure of how “messy” some collection of data is

i = some data k = class index c = total number of classes p(k|i) = probability of randomly selecting item of class k from data

slide-7
SLIDE 7
  • Ex. Gini Impurity

Let’s calculate the Gini Impurity for these groups of data, where the two possible classes are blue or red:

slide-8
SLIDE 8
slide-9
SLIDE 9
slide-10
SLIDE 10
slide-11
SLIDE 11
slide-12
SLIDE 12
slide-13
SLIDE 13
slide-14
SLIDE 14
  • Ex. Gini Impurity

0.5 0.444

slide-15
SLIDE 15
  • Ex. Gini Impurity

0.5 0.444 Minimum possible impurity Maximum possible impurity

slide-16
SLIDE 16

Information Gain

  • Dp , Dleft, Dright are the parent node, left node dataset, and right node dataset respectively
  • I is a measure of impurity (like Gini Impurity)
  • Np, Nleft, and Nright are the number of items in the parent, left, and right nodes respectively
  • f is the question you are asking to create the split
slide-17
SLIDE 17

T B T B B T Age > 27? N T B B Y B T T T B T B B T Height > 6’4’’ N T T Y B B B T T = Tennis Player B = Basketball Player Let’s figure out which question is a better question to ask to split the athletes according to sport

slide-18
SLIDE 18

T B T B B T Age > 27? N T B B Y B T T

slide-19
SLIDE 19

T B T B B T Age > 27? N T B B Y B T T 1/2 4/9 4/9

slide-20
SLIDE 20

T B T B B T Age > 27? N T B B Y B T T 1/2 4/9 4/9

slide-21
SLIDE 21

T B T B B T Height > 6’4’’ N T T Y B B B T

slide-22
SLIDE 22

T B T B B T Height > 6’4’’ N T T Y B B B T 1/2 3/8

slide-23
SLIDE 23

T B T B B T Height > 6’4’’ N T T Y B B B T 1/2 3/8

slide-24
SLIDE 24

T B T B B T Age > 27? N T B B Y B T T T B T B B T Height > 6’4’’ N T T Y B B B T Information Gain: 0.055556 Information Gain: 0.25 Since Information Gain is higher, this the better question to ask to classify our athletes

slide-25
SLIDE 25

How to Come Up with Values for the Questions?

  • The most straightforward way: Try out different values from the items in your training dataset
slide-26
SLIDE 26

Overfitting

  • Techniques to prevent overfitting in decision trees:
  • Continue recursively generating nodes only if information gain is larger than some threshold (e.g.

0.1)

  • After creating the tree, prune all nodes that are at a depth greater than some threshold