The “deep” in Deep Learning refers to the circumstance that we derive the output for our input by passing it through multiple layers / a hierarchy of transformations (i.e., we go through some depth of transforms), instead of having a single, linear formula that tells us the output value directly.