ReLU6 in ML

What's an activation function?

When using machine learning, activation functions are used to transform weighted input from a node to an output value that is passed on to either another layer or can be used as the final output. If we neglected using these, we would simply see linear regression.

A good comparison is thinking about how the brain works. A neuron will get many input signals and its 'activation function' will determine its importance. This activation function would determine whether o...