Alternative Classification Techniques

Question 1

6 ALTERNATIVE CLASSIFICATION TECHNIQUE

Answer 1

1. Rule-Based Classifier
2. Nearest Neighbor Classifier
3. Naïve Bayes Classifier
4. Artificial Neural Network
5. Support Vector Machines
6. Ensemble Method

Question 2

classify record by using a collection of “if…then..” rules.

Answer 2

Rule-Based Classifier

Question 3

These rules are ranked according to their priority.

Answer 3

Ordered Rule Set

Question 4

an ordered rule set is known as this.

Answer 4

Decision List

Question 5

2 RULE ORDERING SCHEMES

Answer 5

1. Rule-based Ordering
2. Class-based Ordering

Question 6

an ordering scheme where individual rules are ranked based on their quality.

Answer 6

Rule-based Ordering

Question 7

an ordering scheme where rules that belong to the same class appear together.

Answer 7

Class-based Ordering

Question 8

fractions of records that justify the antecedent of a rule.

Answer 8

Coverage of a Rule

Question 9

fraction of records that satisfy both the antecedent and consequent of a rule.

Answer 9

Accuracy of a Rule

Question 10

2 Characteristic of Rule-Based Classifier

Answer 10

• Mutually Exclusive Rules
• Exhaustive Rules

Question 11

2 Effect of Rule Simplification:

Answer 11

• Rules are no longer mutually exclusive
• Rules are no longer exhaustive

Question 12

5 Advantages of Rule-Based Classifier

Answer 12

1. Highly expressive as a decision tree
2. Easy to interpret
3. Easy to generate
4. Can classify new instances rapidly
5. Performance comparable to decision tree

Question 13

are lazy learners, it does not build model explicitly, needs to store all training data, and classifying unknown records are expensive.

Answer 13

Nearest Neighbor Classifier

Question 14

3 Requirements for Nearest Neighbor Classifier

Answer 14

1. Set of stored records
2. Distance Metric to compute distance between records
3. The value of k, the number of nearest neighbors to retrieve.

Question 15

3 Ways to Identify Unknown Values in Nearest Neighbor Classifier

Answer 15

1.Compute distance to other records

2. Identify k nearest neighbor
3. Use class label of nearest neighbor (majority vote)

Question 16

is robust to isolated noise points, handles missing valued by ignoring instance during probability estimate calculation, and robust to irrelevant attributes.

Answer 16

Naïve Bayes Classifier

Question 17

involves learning the weights of the neurons.

Answer 17

Artificial Neural Networks

Question 18

Algorithm for learning Artificial Neural Networks (3):

Answer 18

1. Initialize the weights
2. Adjust the weights in such a way that the output of ANN is consistent with class labels of training examples.
3. Find weights that minimize the above function.

Question 19

uses a hyperplane (decision boundary) to separate the data.

Answer 19

Support Vector Machines

Question 20

In SVM, means it is more robust and is less expected to generalization error.

Answer 20

Larger Margins

Question 21

2 Methods for Ensemble Method

Answer 21

1. Construct a set of (possible weak) classifiers from the training data.
2. Predict class label of previously unseen records by aggregating predictions made by multiple classifiers.

Question 22

3 Advantages of Ensemble Method

Answer 22

• Improve stability and often also accuracy of classifiers
• Reduce variance in prediction
• Reduces overfitting

Question 23

What is the general idea of Ensemble Method

Answer 23

1. Create multiple data sets
2. Build multiple classifiers
3. Combine classifiers

Question 24

3 Examples of Ensemble Method

Answer 24

1. Bagging
2. Boosting
3. Random Forests

Question 25

3 Steps of Bagging

Answer 25

• Sampling with replacement (bootstrap sampling)
• Build Classifier
• Aggregate the classifiers’ results by averaging or voting.

Question 26

an example of Ensemble Method where records that are incorrectly classified in one round will have their weights increased in the next.

Answer 26

Boosting

Question 27

is a popular algorithm that typically used a decision tree as the weak learner.

Answer 27

AdaBoost

Question 28

an example of Ensemble Method that introduces two sources of randomness: Bagging and Random Input Vector.

Answer 28

Random Forests

Question 29

each tree is grown using a bootstrap sample of training data.

Answer 29

Bagging Method

Question 30

at each node, best split is chosen only from a random sample.

Answer 30

Random Vector Method