Artificial Intelligence Viva Questions (Mumbai University)

So here is the list of viva questions which were asked:

1. Why did you choose this subject?
2. What do you mean by fuzzy?How is it different from the normal crisp logic?Give example for the         same.
3. What is the algorithm for:
 a.Hebb Rule.
 b.Perceptron
 c.Adaline
 d.Madaline
 e.BackPropagation
4.Which one is the best method?Why?
5.What are the advantages of backpropogation over perceptron;perceptron over backpropogation;
   madaline over adaline & backpropogation over madaline?
6. Design a Fuzzy controller with 2 inputs and one output?
7. List the various defuzzification techniques?Which one is the best & why?
8. Difference between Supervised & Unsupervised Learning?
9. How would you use AI in your project?(this question was asked to almost everyone)
10.What are the steps involved in character recognition?
-} first you need to have a set of images for each character.All the images should be of same                    dimension.Since the images which you have are of RGB(3 dimension)format convert it to gray            scale(2 dimesion). Then convert the 2 dimesional matrix into a single column matrix which is to          be given to the neural network for training.The training algorithm which you use depends on the          application(like if you want an accurate result you can use Backpropagation & if you want a fast        output use Perceptron)