Exam 3

Question 1

What is Unsupervised learning (clustering)?

Answer 1

• the class labels of training data are unknown
• given a set of measurements, observations, etc. with the aim of establishing the existence of classes or clusters in the data

Question 2

What do decision trees do?

Answer 2

identify ways to split a data set

Question 3

What does a decision tree start with?

Answer 3

Root Node

Question 4

What predicts discrete labels?

Answer 4

classification

Question 5

What predicts continuous quantity or values?

Answer 5

regression

Question 6

What does multi-class classification require?

Answer 6

requires that a sample only have one class

Question 7

What is a small portion of a decision tree called?

Answer 7

sub-tree

Question 8

Type of classification algorithms in machine learning? (4)

Answer 8

• linear classifiers
  - k-nearest-neighbors
• decision trees
• support vector machines
• neural networks

Question 9

The data used to view a classification model is called…

Answer 9

Training Data

Question 10

In supervised learning, training data includes both ____ and _____

Answer 10

input & desired output

Question 11

Validation data is used for…

Answer 11

testing the model

Question 12

For SVM the trick is to do ____ ______ data mapping

Answer 12

high dimensional

Question 13

The effectiveness of SVM depends on…

Answer 13

• section
• parameters

Question 14

SVM are useful alternative to which model?

Answer 14

ANN

Question 15

To divide the data into distinct groups so that points in a group are very similar is the main point of what model?

Answer 15

K means clustering

Question 16

Example of non-probabilistic binary linear classifiers

Answer 16

SVM specifically using the kernel method

Question 17

In supervised learning, training data is accompanied by…

Answer 17

class labels indicating the class of observation

Question 18

The mathematical methods of choosing the best split are… (2)

Answer 18

Entropy & Information Gain

Question 19

For decision tree, the splitting method is by…

Answer 19

reduction in variance

Question 20

What is Overfitting?

Answer 20

Model is too specific to training data and may have poor accuracy for unseen samples

Question 21

Two approaches to avoid overfitting

Answer 21

pre-pruning & post-pruning

Question 22

The basic algorithm for decision trees is

Answer 22

recursive partitioning (top-down recursive divide-and-conquer manner)

Question 23

Typically the ______ between each pair of adjacent values is considered as a possible split point

Answer 23

midpoint

Question 24

Random forest used the ____ ____ to construct decision trees

Answer 24

gini index

Question 25

Trees represent knowledge in the form of _________ rules

Answer 25

IF-THEN

Question 26

The motivation for SVM is to categorize new unseen objects into two separate groups based on their ______ and _______

Answer 26

Properties & a Set of Known Examples already categorized

Question 27

What is one of the key areas in machine learning?

Answer 27

Kernel Methods

Question 28

What are the two key concepts of SVM?

Answer 28

• maximize the margin
• the kernel trick

Question 29

What are supervised learning models of associated learning algorithms that analyze data and recognize patterns?

Answer 29

Support Vector Machines (SVM)

Question 30

How do you choose the best support vector in SVM?

Answer 30

Choose the hyperplane that maximizes the margin between classes

Question 31

What are the vectors points that the margin lines touch known as?

Answer 31

Support Vectors

Question 32

Large value of parameter C = _____ margin

Answer 32

Small

Question 33

Small value of parameter C = _____ margin

Answer 33

Large

Question 34

How is distance measured for KNN?

Answer 34

Euclidean distance

Question 35

What do the KNN algorithm assume?

Answer 35

similar things exist in close proximity

Question 36

What is the K value in KNN?

Answer 36

K is the number of existing data points that will be compared to the new data point

Question 37

How are data points assigned in KNN?

Answer 37

The closest “K” neighbors are compared to the new point and assigned to the category in the majority among the neighbors

Question 38

what happens when K is too small?

Answer 38

could be sensitive to noise

Question 39

What happens if K is too large?

Answer 39

neighborhood might include points from other classes

Question 40

The value of k must be : even or odd?

Answer 40

odd to eliminate ties

Question 41

Which model makes NO ASSUMPTIONS about the data?

Answer 41

KNN

Question 42

Typically choose the value of k which has the lowest ____ _____ in _____ data

Answer 42

error rate; validation

Question 43

When using KNN for prediction, the model uses the….

Answer 43

average of response values