Artificial Neural Network 2018 Question Paper — Government College of…
The 2018 Artificial Neural Network previous year question paper (Government College of Engineering and Technology, Jammu), with every question and its marks.
- University: Government College of Engineering and Technology, Jammu
- Subject: Artificial Neural Network
- Course: Artificial Neural Network
- Branch: ELEC. & COMM. ENGINEERING
- Year: 2018
- Paper: end-term
- Duration: 180 minutes
- Total marks: 100
Instructions
- Attempt five question in all, selecting at least two questions from each section.
Questions
Section-A
- Q1(a). Draw and explain the basic model of an artificial neuron. [10 marks]
- Q1(b). What is an activation function ? What are different activation functions in neural networks ? Give examples and explain. [10 marks]
- Q2(a). Describe Delta learning rule. How Widrow and Hoff LMS learning rule can be treated as a special case of Delta rule. [10 marks]
- Q2(b). What is self-organizing map ? Explain Kohonen's self-organizing map with learning algorithm. [10 marks]
- Q3(a). Draw the architecture of a multi-layer perceptron and explain its operation. Mention its advantages and disadvantages. [10 marks]
- Q3(b). What is clustering and what are different methods of clustering ? Discuss winner takes all learning networks. [10 marks]
- Q4(a). ART Networks. [10 marks]
- Q4(b). RBF. [10 marks]
Section-B
- Q5(a). What is a Linear Association ? What are its limitations as an associative memory ? [10 marks]
- Q5(b). Describe the model of Hopfield Associative Memory. [10 marks]
- Q6(a). If $\mu_R = \{0.3/30 + 0.7/60 + 1.0/100 + 0.2/120\}$ and $\mu_S = \{0.2/20 + 0.4/40 + 0.6/60 + 0.8/80 + 1.0/100 + 0.1/120\}$ (i) Find the relation (T) between R and S using Cartesian product. (ii) If $\mu_N = \{0.33/500 + 0.67/1000 + 1.0/1500 + 0.15/1800\}$, find another relation (Q) between S and N using Cartesian product. (iii) Find $Z = T \circ Q$ using max-min composition. [15 marks]
- Q6(b). Explain the difference between a classical set and a fuzzy set. [5 marks]
- Q7. What is defuzzification ? Explain different methods of defuzzification in detail. Which method is the most efficient one ? [20 marks]
- Q8(a). Basic fuzzy inference algorithm [10 marks]
- Q8(b). Properties of fuzzy set. [10 marks]