# Forward Propagation Forward propagation is the process through which data passes through layers of neurons in a [[Neural Networks (NNs)]]. ![[Forward Propagation.png]] The data flows through each neuron by connections or "dendrites". Every connection has a specific weight by which the flow of data is regulated. In the example above, `x1` and `x2` are the two inputs. They could be integers of floats. When those inputs pass through the connections, they're adjusted depending on the weights, `w1` and `w2`. The neuron then processes the information by outputting a weighted sum of these inputs. It also adds a constant to the sum which is referred to as the *bias*. `a` is the output of the network. Forward propagation is much more interesting when it maps the weighted sum of the inputs to a nonlinear space. The [[Sigmoid Function]] ([[Activation Functions]]) does that. # Conventions - `z` always represents the linear combination of the inputs - `a` always represents the output of a neuron