‘Siri, tell me if it’s going to rain today.’ ‘Facebook, tag my friend in this photo.’
These are just two examples of the things we ask computers to do for us. But have you ever wondered how they’re able to do it?
Machine learning is not a new concept, but it is constantly evolving, and the potential benefits of its capability are increasing by the second. A form of artificial intelligence, it provides computers with the ability to learn through experience, without being explicitly programmed to perform a task. As the computer receives more data, its algorithms become more finely tuned, and over time it begins to recognise patterns and solve problems on its own – without the use of a programme. The more finely tuned the algorithm, the more accurate the computer can be in its predictions.
Through an animation featuring machine learning expert Professor Yee Whye Teh of the Department of Statistics, Oxford Sparks – the University of Oxford’s digital science portal – demonstrates how researchers have combined the power of statistics and computer science to build algorithms capable of solving complex problems more efficiently while using less computer power.
Using machine learning in this way is already informing medical diagnosis and strengthening the speed and capability of smartphones and social media – but its scope to revolutionise the world seems limitless.