Consider a two-input neuron with weight vector $W = [3, 2]$, input vector $X = [-5, 7]^T$ and bias $b = 1$. Which of the following transfer or activation functions can be used to produce an output of 0.5?
- Artificial Neural Network Neuron Activation:
- Given neuron parameters:
- Weight vector: $W = [w_1, w_2] = [3, 2]$
- Input vector: $X = [x_1, x_2]^T = [-5, 7]^T$
- Bias: $b = 1$
- Calculating net input ($n$):
- Evaluating transfer functions at net input $n = 0$:
- Log-Sigmoid (Logistic Function):
- Hard limit: Returns 1 for $n \ge 0$, and 0 for $n \lt 0$.
- Linear: Returns $f(0) = 0$.
- Conclusion: Log-sigmoid transfer function produces an output of 0.5 (Option 4).
Correct option: (4) Log-sigmoid