Tut 3

Question 1

Annoying thing to remember with sequential delta learning algorithm (LTU)

Answer 1

Because formula for linear threshold neuron is
Y = H(wx - θ)
We need to initialise w for augmented vector to be
[-θ, w1, w2]

Question 2

Thing to remember about negative feedback networks

Answer 2

(Type of competitive learning network)

2 separate phases, may get asked to only do several activation steps

Question 3

Neuron def

Answer 3

An information processing unit

Question 4

Def action potential

Answer 4

The signal outputted by a biological neuron

Question 5

Firing rate

Answer 5

The number of action potentials emitted during a defined time period