Perceptron network with step(heaviside) activation signal

Hello friends, how are you?

I’m starting my studies in ANN and I would like to make a perceptron network with the activation signal heaviside(step).

Unfortunately I couldn’t find anything on the internet, could anyone help me?

The dataset I will use:
input = torch.tensor([[-1.0,0.1, 0.4, 0.7],
[-1.0,0.3, 0.7, 0.2],
[-1.0,0.6, 0.9, 0.8],
[-1.0,0.5, 0.7, 0.1]])
values = torch.tensor([1.0, -1.0, -1.0, 1.0])

Thank you!

The Neural Networks tutorial might be a good starter. The heaviside function can be found here but note that it’s not (meaningfully) differentiable as it would yield zero gradients almost everywhere.