Detlef W.M. Hofmann is senior researcher in the group of Biosciences at the Center for Advanced Studies, Research, and Development in Sardinia (CRS4), Italy since 2005. He holds a doctor degree (1991) in Theoretical Chemistry at the University of Erlangen/Nuremberg in the Dept. of theoretical chemistry under Prof. J. Ladik and Prof. P. van Schleyer. At the beginning of his scientific carrier he worked together with world famous scientists such as J. Ladik, G. Del Re, E. Clementi, and T. Lengauer, who highly estimated his abilities and helped to look on problem from the different point of views and to provide an open view to new fields. Nowadays, Detlef W.M. Hofmann is a leading scientist in the field of information retrieval from data bases of computational chemistry and crystallography.
He belongs to the first researchers, who proposed 25 years ago to use machine learning in chemistry. He has adapted different methods of machine learning on experimental data bases. Typical examples are clustering to recognize similarity between crystals and molecules, multidimensional regression to generate increment systems for instance for the density prediction of crystals, anomaly detection to find false publications, dimensionality reduction to simplify formula, and supervised learning to derive the Gibbs energy. The Gibbs energy is the crucial variable for all chemical reactions and its calculation opens a wide field of predictions. Till now it was used to predict crystal structures, isotopic effects, phase transitions, and co-crystal formation.