Thapar University 2007 B.E PS-007 INTELLIGENT TECHNIQUES & APPLICATIONS - Question Paper

THAPAR UNIVERSITY PATIALA
ELECTRICAL & INSTRUMENTATION ENGINEERING DEPARTMENT
MID TEST 2

THAPAR UNIVERSITY, PATIALA Electrical & Instrumentation Engineering Department Mid-Test II

PS-007 (Intelligent Techniques & Applications)

M.M. : 30    Tim : On Hour

Q.1 (a) Explain why the learning rate is kept small positive in Neural Network (2)

(b) List the salient features of recurrent neural network. Also draw an schematic arrangement    (4)

Q.2 Discuss the Newtons method for unconstrained optimization leading to updation of weights iteratively    (8)

Q.3 With the help of an appropriate single loop feedback system show that weight |w|<1 corresponds to stable convergent system. Simultaneously, it corresponds to system of infinite fading memory.    (8)

Q.4 Implement the logical XOR function using McCulloch-Pitts neuron model (8)

Attachment: thapar-university-2007-b-e-25946-1.bmp

Earning:   Approval pending.