Population-based training (PBT) falls primarily into the categories of artificial intelligence, big data, smart data and digital society. This term refers to a method used to improve artificial intelligence (AI) by training algorithms based on a large amount of data from the entire population and not just on a small, selected data set.
Imagine an AI designed to support medical diagnoses. If it is only trained with data from a small group of people, it could deliver incorrect results for other population groups. In population-based training (PBT), algorithms are trained with very large and diverse data sets - for example with anonymised health data from millions of people of different origins, age groups and regions.
This increases the accuracy and reliability of AI models in everyday life. PBT therefore ensures that digital applications work better for everyone - whether it's personalised medicine, traffic control or smart household appliances. This makes artificial intelligence fairer and more usable because it is based on real data from society.















