This preview shows page 1. Sign up to view the full content.
Unformatted text preview: . List all the other distinct possibilities for C1. Note: two clauses are distinct from each other if they are not logically equivalent.] (a) (3) Consider the class of neural nets with inputs x1 : : :xn , which are either 0 or 1. A network is speci ed by giving the weights on the links and the activation function g at each node. Specify a network that computes the majority function for 5 input nodes. That is, it should output 1 if at least half the inputs are high, and 0 otherwise. (b) (3) Draw a decision tree that represents the disjunction of ve inputs. (c) (4) Describe why and how you might apply simulated annealing to train a neural network. (d) (4) Suppose you are training a neural network in a genuinely nondeterministic domain. The training set consists of N copies of the same example, a fraction p > 0:5 of which are positive and a fraction 1 ; p of which are negat...
View Full Document
- Spring '08
- Computer Science