How do algorithms learn? How can we define learning? As humans, we learn a lot throughout our lives. It is a natural task for us. In the first few years of our lives, we learn how to control our body, walk, speak, and recognize different objects. We constantly get new experiences, and these experiences change the way we think, behave, and act. Can a piece of computer code learn like we do? To approach machine learning, we first need to come up with a way to transmit experience directly to the algorithm.
In practical cases, we are interested in teaching algorithms to perform all kinds of specific tasks faster, better, and more reliably that we can do ourselves. For now, we will focus on prediction and recognition tasks. Thus, we want to build algorithms that are able to recognize patterns and predict future outcomes. The following table shows some examples...