Machine Learning algorithms learn to perform a task autonomously or to make predictions from data. Unlike a traditional computer programme, a Machine Learning system does not follow instructions, but learns from experience. Its performance improves as it is “trained” and the algorithm is exposed to more data.