What is Machine Learning?
Machine learning (or ML) has been a buzzword for years - but how and what does a machine learn?
Machine learning, not trying to be rude, is exactly what it says on the tin - when machines learn something (by themselves, without being told exactly what to do by humans).
IBM defines Machine Learning as "a branch of artificial intelligence (AI) and computer science which focuses on the use of data and algorithms to imitate the way that humans learn, gradually improving its accuracy."
The Three Stages of Machine Learning
Looking at the diagram above, we can see three distinct stages of the Machine Learning programme. These three stages were defined by UC Berkeley which we have summarised as:
1. The decision or prediction
2. Error evaluation
3. Optimising itself
A Relatable Example - Weather Predicition
Let's look at a very simplistic example of a computer programme used to predict the weather. A non-ML programme would look at previous weather data (and maybe some known trends or logic, e.g. if it is raining at 1pm there is a 80% chance it will be raining at 2pm). An ML programme would also use a similar method, but it would also look at whether or not its prediction was correct (i.e. was it raining at 2pm?), and then use this data to refine or update future predictions. In the non-ML programme, a human would have to look at the real world data and then make decisions to update the programme in the future. In the ML programme, this would be automatic.
Types of Machine Learning - Supervised learning, unsupervised learning and reinforcement learning
Examples of Machine Learning Algorithms - recommender systems, decision trees, neural networks
Are ML and AI the same thing?