Tut 1

Question 1

Exemplar

Answer 1

A particular datapoint which is represented by a feature vector

Question 2

Dataset

Answer 2

The collection of feature vectors for all exemplars

Question 3

Generalisation

Answer 3

How well a model performs on new data

Question 4

Overfitt8ng

Answer 4

Making the model so specific to the training data that it fails to generalise to new data

Question 5

Decision theory

Answer 5

Methods of making decisions that reduce cost rather than misclassification rate

Question 6

Linearly separable

Answer 6

Exemplars from 2 classes can be separated by a hyperplane in feature space

Question 7

Grid search

Answer 7

A method of trying to find suitable le hyperparameters that searches all possible combinations of values within given ranges

Question 8

Training data

Answer 8

The collection of exemplars used by the learning algorithm to tune parameters of a classifier

Question 9

Define regression

Answer 9

A method that learns to predict a continuous value for each exemplar

Question 10

F1 score

Answer 10

(2 × Recall × Precision)/(recall + precision)