Perceptron: iterating weights

[GB449]

In the perceptron model where we iterate the weights, w(t+1) = w(t) + y(t)x(t), w0 (bias), will be an integer for all t, if we start with zero weights, correct? Since each x(t) = 1 for all t and all x.

[magdon]

Yes, if you start the weights at 0, the bias will be an integer since .