Researchers at the University of Hamburg have created an artificial intelligence process marking the first use of artificial intelligence in quantum physics, the “Nature Physics” journal reported Monday (July 1, 2019). The method improves the identification of quantum transitions i.e. points at which the properties of substances change from experimental data. Artificial intelligence and machine learning can be applied in diverse fields ranging from autonomous driving to fully automated industrial processes.
Simplified measurement and data analysis
The measurement of quantum phase transitions using conventional evaluation methods is far more time-consuming than the simplified analysis of large amounts of data by machine learning. The results can have far-reaching consequences for everyday research, the physicists stressed. Artificial intelligence can analyse entirely new effects of quantum physics in the laboratory in real time, which are otherwise inaccessible. “The application of machine learning techniques to quantum gas experiments opens up many exciting possibilities,” said Niklas Käming, a Master’s student involved in the analysis of the data. The method will be expanded to so-called unsupervised machine learning in the next step.