Ability of a machine to use algorithms to analyse their environment, learn from data and use what has been learnt to take actions and make decisions – with some degree of autonomy – to achieve specific goals.
Comment
- An AI system can learn from its environment by producing an output (predictions, recommendations or decisions) for a given set of factors or feature variables;
- there are three main strands of artificial intelligence:
- machine learning: application of AI that provides machines with the ability to learn automatically and improve from experience (training dataset) without being explicitly programmed;
- deep learning: AI technique where artificial neural networks with various layers, algorithms based on the structure and functioning of the human brain, learn from large amounts of data to create patterns for decision-making;
- reinforcement learning: computational approach to experiential learning whereby an autonomous agent (robot, etc.) tries to maximise the total amount of ‘reward’ it receives over time when interacting with a complex, uncertain environment.
Source
Based on Rouhiainen, 2018; OECD.AI, 2021; expert.ai; The Economist, 2017; Sutton R. and Barto, A., 1998.