NN 1

5 flashcards · Computer Science · by J M COOPER

1. Why do we need a hidden layer to model XOR? Draw a NN for XOR.

   With inputs going straight to the output layer, we can only model situations where there is a simple, separable split between output values. There is no simple threshold or combination of weights that will model XOR without an additional layer.

2. Write as an equation:

3. What is an activation function?

   A simple threshold function at each interior and output node (not normally the same function)

4. What is a bias term?

   Bias terms in NN are equivalent to intercept terms in regression Suppose we want to model x, y data with a straight line The basic equation for a straight line is y = mx However, all these lines go through (0, 0) The equation for any straight line is y = mx + c where c is the intercept term.

5. What is a neural network and how may it be represented?

   Defined by inputs, layers, nodes, weights, activation function and output (scaling) function Inputs feed forward to the output layer At each stage, bias terms may be added, weighted sums are calculated and used as inputs to the activation or output functions The NN structure can be represented by a chart or an equation of the form y = f(W, x) where y is the output (which could have multiple values), x is a vector of numerical inputs, W are the weights, and f is the overall function that determi…