The more i learn about machine learning, the more i see as being a mechanical process. There are two school of thoughts in the computing world, one that says that computing is a field of it’s own and one that says that it’s an expression of maths.
Today, the gold rush of the industry is Ai. Ai is a vast term but, when companies say Ai today, they mean machine/deep learning. The core of Ml/Dl is maths. They essentially have a formula which they tune according to the training data to output a result then use it to compare the result of new samples. Or, they learn patterns and output variations of it.
It is driven by numbers, using a variety of techniques. And probability, a lot of it. The goal of computing has been automation. Automating calculations is the core of it. The recent variation of Ai has been a sublime illustration that computing is indeed maths.
This might seem weird but taking advanced ml/dl courses at the end of the day boils down to learning and understanding formulas by deriving them from scratch. It also includes knowing how to use a tool/library to do so. Also about techniques to patch results we don’t want. It’s very boring, it’s dry maths at work.