Unsupervised representation learning is a term from artificial intelligence and is frequently used in the field of big data and smart data. It concerns how machines or computer programs independently recognise patterns and correlations from large amounts of data – entirely without humans specifying what they should pay attention to.
Unlike supervised learning, where a human demonstrates the correct answers, unsupervised representation learning works independently. For example, an artificial intelligence can evaluate millions of images and automatically discover commonalities within them, such as that many pictures of dogs have similar features. In this way, the AI learns what a „dog shape“ looks like without ever being explicitly told what a dog is.
This procedure helps companies to recognise hidden structures in their data – for example, to identify customer groups, discover trends or align products better with user needs. Especially in times of ever-increasing amounts of data, unsupervised representation learning is an important tool for digital transformation and innovation.
Deutsch 
English (UK)