Types Of ML

We now understand that computer models can learn from training data.

But that doesn’t answer the question – how?

We’ll get into some specifics soon, but first, there are few broad types of machine learning to understand:

• Supervised Learning — Here, data is "labelled", letting the model compare its predictions to correct answers.
• Unsupervised Learning — Data is unlabelled, meaning the model is left to find its own patterns and draw its own conclusions.
• Self-Supervised Learning — The model labels its own data (thus "supervising" itself).
• Reinforcement Learning — The model is trained using a system of "reward" and "punishment", much like a silicone dog made of mathematics.

Much like in humans, spending too much time memorising arbitrary data has its downsides.

So, to find the sweet spot between myopic poindexter and unemployable generalist, we need to understand some tradeoffs.