A single perceptron model, which includes a bias weight, uses a hard-limiter as the output activation function, therefore: We can use Gradient Descent to find the weights that minimizes the error. The decision boundary is, in general, a hyperplane. O Both of the above are true O None of the above are true