In an artificial neural network,what must be true about the threshold value of a neuron that processes an output of 1 when all of its inputs are 0?
A) It is less than -2.
B) It is not positive.
C) Both A and B are true.
D) Nothing can be determined without knowing the weights.

Question

Quizplus · Accepted Answer

For a neuron to output 1 when all inputs are 0, the bias (or threshold) must be set in such a way that it activates the neuron. Since weights multiply the inputs (which are 0), the only factor that can activate the neuron is the bias. If the bias is not positive (i.e., it could be zero or negative), it wouldn't be sufficient to activate a typical neuron designed to output 1 under these conditions. Therefore, the statement that the threshold is not positive is incorrect in this context; it must indeed be positive to ensure activation in the absence of input. This correction highlights the importance of the bias/threshold value in activating a neuron independently of its inputs.

In an Artificial Neural Network,what Must Be True About the Threshold