The term Machine Learning gets thrown around a lot recently, but what does it really mean and how does it work? Imagine a task that is simple for us humans, reading a sentence, or even just recognizing a character on a piece of paper. That is a task so simple that it is literally taught to 5-year-olds, some of which are not yet trained how to use a toilet, but for a computer this is normally a very difficult task. A program is a collection of statements and conditions, meaning that the programmer tells a computer what to do for every situation it encounters. This is not possible for things like handwriting; everyone writes differently. So how can machines read handwriting? That is where machine learning comes in.
Machine learning, or ML, is the process of giving a program data and it learns from the data what it should do. One of the most popular forms of machine learning, Supervised Learning, takes a set of training data both the inputs and the outputs, and the program attempts to learn what it needs to do to the input in order to produce the output. So imagine showing a computer through a camera all the letters of the alphabet and telling it what letter each symbol represents. After repeating this a few hundred or thousand or million times depending on the program, which is a lot faster for a computer to do than for us, the computer learns what kinds of symbols represent what letters. Then it takes new inputs and predicts what the output should be based on how it was trained. Bam, you have a machine that learned how to read.
This is a very exciting new field in the realm of artificial intelligence, and there are many layers of machine learning that I didn't talk about here: Neural networks, evolutionary algorithms, deep learning, reinforcement learning, etc. All different ways to do essentially the same thing described here. If you want to learn how to make programs that learn, come and take UAT's Artificial Intelligence program.
keywords: machine learning, artificial intelligence, supervised learning, unsupervised learning, reinforcement learning, neural networks, evolutionary algorithms, deep learning, programming, computer science