It’s a tool whereby we can program things that might require what previously required human intelligence. It ties deeply into statistics.
In various points of it’s history, it was also known as:
- predictive analytics
- self-teaching computers
It has two main concerns:
- classification of data based on some model
- make predictions for the future based on these classifications
As a concrete example, we may have photos of freckles. If we can classify them into cancerous and non-cancerous, we could then use that information on predictions for whether or not the next photo presented is cancerous or not.