Input
Remember how a biological neuron takes in electrical impulses from its dendrites? Well, the perceptron behaves in a similar fashion, yet it prefers to ingest numbers in lieu of electricity. Essentially, it takes in feature inputs, as shown in the preceding diagram. This particular perceptron only has three input channels, these being x1, x2, and x3. These feature inputs (x1, x2, and x3) can be any independent variable that you choose to represent your observation by. Simply speaking, if we want to predict whether it will be sunny or rainy on any given day, we can record independent variables such as temperature and air pressure per day, along with the appropriate output class of that day (whether the day itself was sunny or rainy). We will then feed these independent variables that we have, one day at a time, as input features into our perceptron model.