A typical Perceptron neural network application is

classification. Consider the simple example of classifying trucks given

their masses and length, with targets t1, t2, t3 and t4, assume a two-

classes problem: (a) Construct a Perceptron neural network that can

classify any truck to either class 1 (Lorry class) or class 2 (Van class). The

Perceptron network is set to have 2 neurons (S=2), where a Hardlims

Transfer functions are used. Initial weights are set to a matrix of zeros,

while the biases of all neurons are set to -1. Givena 4 prototype input

vectors of the learning set (as shown in the table above), apply only one

iteration of the Perceptron Learning Rule to fine the output of the

Perceptron Network and calculate the learning Error.(b) Write a MATLAB

CODE (or Python) to build a perceptron Neural Network, then Train using

the above 4 learning Input prototype vectors using a = 0.5, log sigmoid

transfer function and convergence Error threshold: 10-3.

Mas

W

Mass

Length

Class

a1. 1)

P1

Class 1

s00

2=(1.1)

Pe

Class 1

200

5

w.

a2

P

3=(0,0

Classs

W

4(0.0)

Class s

2

P#

Length

8 points Question 3: A typical Perceptron neural network application is classification. Consider the simple example of classifying trucks given their masses and length, with targets t1, t2, t3 and t4. assume a two- classes problem: (a) Construct a Perceptron neural network that can classify any truck to either class 1 (Lorry class) or class 2 (Van class). The Perceptron network is set to have 2 neurons (S=2), where a Hardlims Transfer functions are used. Initial weights are set to a matrix of zeros, while the biases of all neurons are set to -1. Given a 4 prototype input vectors of the learning set (as shown in the table above), apply only one iteration of the Perceptron Learning Rule to fine the output of the Perceptron Network and calculate the learning Error.(b) Write a MATLAB CODE (or Python) to build a perceptron Neural Network, then Train using the above 4 learning Input prototype vectors using a = 0.5, log sigmoid transfer function and convergence Error threshold: 10-3. WA Mass Length Class Ma33 0 w.. 100 6 Class 1 20,0 5 Class 1 wa Pio Pe PSD P4 > H = (1, 1 12 = (1,1) 13 = (0,0) 14 = (0,0) -O2 5.0 4 Classe W B w 20 5 Class W.. Length o