Chapter 1- Introduction

Question 1

what is machine learning?

Answer 1

Programming a computer so that it can learn from experience

Question 2

what does it mean to “learn”?

Answer 2

adaptation in response to observed data

Question 3

what is supervised learning?

Answer 3

Ground truth labels are provided

Question 4

what is unsupervised learning?

Answer 4

Ground truth labels are unknown

Question 5

what is overfitting?

Answer 5

Model learns too well on training data and cannot predict well on testing data

Question 6

what does it mean to say some ML algorithms are stochastic?

Answer 6

They rely on random initialisation of some parameters

Question 7

what are the two learning paradigms in machine learning

Answer 7

deterministic

probabilistic (or stochastic)

Question 8

describe deterministic ML

Answer 8

one output for one input and it is always the same

Question 9

describe probabilistic (or stochastic) ML

Answer 9

One input can have many outputs. Given the same conditions and inputs a learning algorithm returns a distribution of outputs

Question 10

in probabilistic ML, the function f in X => Fp(Y) is…

Answer 10

a stochastic process

Question 11

3 reasons probability is necessary for machine learning:

Answer 11

Finite training data

Data uncertainty

Prediction uncertainty