The term "informed ML models" originates from the fields of artificial intelligence, big data and smart data as well as Industry and Factory 4.0. These are machine learning models (ML stands for machine learning) that not only learn from data, but also utilise additional knowledge or expert information. Such knowledge can come from manuals, expert statements or existing research results, for example.
This additional information helps the ML model to make better and more accurate predictions because it does not start "from scratch" when analysing. Instead of just learning from huge amounts of data, relevant knowledge is incorporated in a targeted manner.
An illustrative example: In a factory, a machine is supposed to use machine learning to predict when a defect might occur. Instead of relying solely on the collected machine data, the model is also fed the expertise of technicians. For example, they know that certain noises often indicate bearing damage. With this expert knowledge, the ML model recognises potential problems earlier and more accurately.
Informed ML models are particularly useful when a lot of data is available, but additional specialised knowledge makes the decisive difference. They combine the best of man and machine.
Deutsch 
English (UK)