Intro

Question 1

Embedded agent

Answer 1

computer systems that sense and act on their environments, monitor complex dynamic conditions and affect the environment in goal-directed ways

Question 2

Neural network

Answer 2

a machine that is designed to model the way in which brain performs a particular task or function of interest

Question 3

Neural network properties

Answer 3

1. Nonlinearity (in case if the mechanism generating input signal e.g. speech signal, is inherently nonlinear);
2. Input-Output Mapping (for supervised learning);
3. Adaptivity (adapt their weights to the changes in the surrounding environment)

Question 4

Learning methods

Answer 4

1. Supervised (given observations and their labels - predict labels of new observations; classification and regression)
2. Unsupervised (given observations - built representation or extract relevant structure; dimensionality reduction, clustering categorisation, course separation)
3. Reinforcement (given states, actions and rewards - find action that maximises reward)