It’s important to make sure we’re all on the same page when it comes to AI, ML, and all the other important acronyms you hear in every conversation about the future of news these days. So we have a one sentence definition of machine learning, as well as a quick intro into the various types of machine learning.
What is machine learning?
Machine learning can often seem to be just another buzzword thrown around in the industry with not everyone on the same page of what exactly it means. So we asked our tech team to write a short explanation to make sure we’re all discussing the same thing.
“Machine learning uses algorithms that gradually improve on a task without explicitly being told how, i.e. they ‘learn’ from data.” – Joris Gielen, AI & Software Engineer at Twipe
Are there different types of machine learning?
There are three main types of machine learning:
- Supervised learning: the machine aims to answer a predefined question
- The spam filter on your email account is an example of classification supervised learning, it learns from the previous emails you’ve marked as spam
- A weather app uses regression supervised learning and historical data about the weather to make predictions about weather
- A churn prediction model uses the characteristics and behaviour of previously churned subscribers to predict which subscribers are likely to churn in the future
- Unsupervised learning: there is no predefined question, rather the machine tries to describe the current environment and explain how customers behave
- Sending targeted marketing to subgroups based on age is an example of clustered unsupervised learning
- Most data visualisation uses dimensionality reduction unsupervised learning to reduce the variables of a data set by finding commonalities
- Reinforcement learning: uses a machine’s own personal history and experiences in order to make decisions
- Machines that can play chess use reinforcement learning to be able to improve to eventually beat human players
For more on how newspapers can use machine learning, make sure to read our guide to what our tech team wants publishers to know about machine learning.