Vous êtes sur la page 1sur 7

www.jntuworld.

com

Code No: 07A7EC01

R07

Set No. 2

IV B.Tech I Semester Examinations November 2010 IV B.Tech I Semester Regular Examinations,NOVEMBER 2010 NEURAL NETWORKS AND FUZZY LOGIC Common to Aeronautical Engineering, Instrumentation And Control Engineering, Electrical And Electronics Engineering Time: 3 hours Max Marks: 80 Answer any FIVE Questions All Questions carry equal marks

1. Explain the following terms : (a) Relation matrix. (b) Binary relation. (c) Identity relation. (d) Universal relation.

2. State and explain the generalized delta learning rule applied in back propagation algorithm. [16] 3. Explain the classication model, features and decision regions in single layer perception. [16] 4. (a) What are the three models of articial neuron? Explain them in detail. (b) Compare and contrast articial neural networks with conventional computer system. [8+8] 5. (a) Dene membership function. (b) Write the components of a fuzzy logic system and explain them. 6. (a) Explain the Hopeld network algorithm and its limitations. (b) Explain the Energy analysis of Discrete Hopeld Network. [8+8] [8+8]

T N

W U

R O

D L

[16]

7. (a) What are invariant characteristics of neuro computing model? Explain each of them. (b) What are variable characteristics of neuro computing model? Explain each of them. [8+8] 8. Write note on the following. (a) Bidirectional Associate memories. (b) Grossberg layer. [8+8]

www.jntuworld.com

www.jntuworld.com

Code No: 07A7EC01

R07

Set No. 4

IV B.Tech I Semester Examinations November 2010 IV B.Tech I Semester Regular Examinations,NOVEMBER 2010 NEURAL NETWORKS AND FUZZY LOGIC Common to Aeronautical Engineering, Instrumentation And Control Engineering, Electrical And Electronics Engineering Time: 3 hours Max Marks: 80 Answer any FIVE Questions All Questions carry equal marks

1. (a) what are the advantages and disadvantages of Neural networks. (b) Explain the applications of Neural networks in Forecasting. 2. Briey explain the Winner take all learning with an example. [8+8] [16]

3. Explain clearly how the perceptron training algorithm is implemented on a digital computer for pattern recognition. [16] 4. (a) With neat diagram explain discrete-time bidirectional associative memory. Also explain encoding and decoding of information in BAM. (b) Draw and explain the architectural graph of a Hopeld network consisting of N=4 neurons. [8+8] 5. For the data shown in the following data table, show the rst 2 iterations in trying to compute the membership values for the input variables R, B and Y rejoins L & R B Y L M M. 0.1 0.02 0.3 1 0 (a) Use a 3 x 2 x 2 x 2 neural network with initializing all the weights to 0.5. (b) Use sigmoid activation function. 6. (a) Explain Partition and Covering with an example. (b) Let the sets A,B,C and E be given as follows E = all students enrolled in the university cricket club=600 A (male students) = 300, B (bowlers)=225 and C (batsmen) =160.let the number of male students who are bowlers (A B) be 100, 25 of whom are batsmen too (A B C), and the total number of male students who are batsmen (A C) be 85.Determine the number of students who are: i. ii. iii. iv. Females. Not bowlers. Not batsmen. Female students who can bowl. [16]

T N

W U

R O

D L

[16]

7. Discuss the operation of single neuron system. A neuron j receives inputs from four other neurons whose activity levels are 10, -20, 4 and -2. The respective synaptic weights of the neuron j are 0.8, 0.2, -1.0, and -0.9. Calculate the output of neuron for the following two situations 2

www.jntuworld.com

www.jntuworld.com

Code No: 07A7EC01 (a) The neuron is linear.

R07

Set No. 4

(b) The neuron is represented by a McCulloch-Pitts model. Assume that the bias applied to the neuron is zero. [16] 8. Using back propagation learning, nd the new weights for the network shown in gure 1, when presented with an input (0,1) and the target output is 1. Use a learning rate of = 0.5 and the binary sigmoid activation function. [16]

T N

W U
Figure 1:

R O

D L

www.jntuworld.com

www.jntuworld.com

Code No: 07A7EC01

R07

Set No. 1

IV IV B.Tech I Semester Examinations November 2010 2010 B.Tech I Semester Regular Examinations,NOVEMBER NEURAL NETWORKS AND FUZZY LOGIC Common to Aeronautical Engineering, Instrumentation And Control Engineering, Electrical And Electronics Engineering Time: 3 hours Max Marks: 80 Answer any FIVE Questions All Questions carry equal marks

1. (a) Dene classical set (b) Dierentiate fuzzy set from classical set and name the properties of classical (crisp) sets. [8+8] 2. (a) Dene Hamming Distance. (c) Which vectors are all called cross talk vectors.

(b) State the application procedure of an auto associator with threshold function.

3. Explain briey about intelligent control. Give few applications in industries. [16] 4. Using your own intuition and your own denitions of the universe of discourse, plot fuzzy membership functions for the following variables : Weight of people (a) Very light. (b) Light. (c) Average. (d) Heavy.

(e) Very heavy.

T N

W U

R O

D L

[16]

[16]

5. (a) Write short note on various learning techniques rate. (b) If the net input to an output neuron is 0.64,calculate its output when the activation function is i. Binary sigmoidal. ii. Bipolar sigmoidal. [10+6]

6. Generalize the XOR problem to a parity problem for N(>2) variables by considering a network for the two variables rst and then extending the network considering the output of the rst network as one variable and the third variable as another. Repeat this for n=4 and design a network for solving the parity problem for 4 variables. [16] 7. For the following noisy versions of training patterns, identify the response of network by saggregating it into correct, incorrect or indenite (0 -1 1), (0 1 -1), (0 0 1), ( 0 0 -1), ( 0 1 0), (1 0 1)

www.jntuworld.com

www.jntuworld.com

Code No: 07A7EC01

R07

Set No. 1

(1 0 -1), (1 -1 0), (1 0 0 ), (1 1 0), ( 0 -1 0), (1 1 1) Say the weights taken from the bipolar step function are W1 = 0, W2 = -2, W3 = 2. [16] 8. Using the perceptron learning rule, nd the weights required to perform the following classications. Vectors (1 1 1 1),(-1 1 -1 -1) and (1 -1 -1 1) are members of class (having value -1). Use learning rate of 1 and starting weights of 0. Using each of the training and vectors as input , test the response of the net. [16]

T N

W U

R O

D L

www.jntuworld.com

www.jntuworld.com

Code No: 07A7EC01

R07

Set No. 3

IV B.Tech I Semester Examinations November 2010 IV B.Tech I Semester Regular Examinations,NOVEMBER 2010 NEURAL NETWORKS AND FUZZY LOGIC Common to Aeronautical Engineering, Instrumentation And Control Engineering, Electrical And Electronics Engineering Time: 3 hours Max Marks: 80 Answer any FIVE Questions All Questions carry equal marks

1. Explain the following terms in brief. (a) Associative memory. (b) Stimuli. (c) Memory convergence. (d) dummy augmentation.

2. (a) Describe the Rosneblatts perceptron model of an articial neuron. (b) Distinguish between linearly separable and linearly unseparable problems. Give examples for each. [8+8] 3. (a) What is the optimization technique used in back propagation algorithm ? (b) State the application algorithm of a back propagation net ? [8+8]

4. Design and develop a pressure process control by Fuzzy Logic Control model. Formulate necessary membership functions and required fuzzy rules for the application. [16] 5. (a) Give and explain the properties of crisp sets (b) Let R,S be dened on the sets {1, 3, 5} {1,3,5}. Let R: {(x, y)/y = x + 2}, S : {(x, y)/x y}/. Using max min composition nd i. RoS ii. SoR. [8+8]

T N

W U

R O

D L

[16]

6. Using your own intuition, develop fuzzy membership functions on the real line for the fuzzy number approximately 2 AND approximately 8, using the following function shapes: (a) Symmetric triangles (b) Trapezoids. (c) Gaussian functions. [16]

7. (a) What is an articial intelligent technique? How is it dierent from the technique of conventional programming? (b) Write briey about the major areas of AI. 6 [8+8]

www.jntuworld.com

www.jntuworld.com

Code No: 07A7EC01

R07

Set No. 3
[8+8]

8. (a) What are the stopping conditions used to stop the progress of the training algorithm. (b) Explain the algorithm used for training the perceptron net.

T N

W U

R O

D L

www.jntuworld.com

Vous aimerez peut-être aussi