From Machine Learning to explainable AI
The Holzinger Group HCI-KDD promotes an synergistic approach of two areas to understand intelligence, enabling context-adaptive systems: Human-Computer Interaction (HCI) & Knowledge Discovery/Data Mining (KDD). The goal is to augment human intelligence with artificial intelligence. The Holzinger Group has pioneered in interactive machine learning (iML) with the human-in-the-loop. Raising legal issues in the European Union (General Data Protection Regulation) make glass box approaches important, i.e. to be able to make decisions transparent, re-traceable thus understandable, to explain why a machine decision has been made – towards explainable-AI.
- To reach a level of usable intelligence we need to learn from prior data, extract knowledge, generalize, fight the curse of dimensionality, disentangle underlying explanatory factors of data, i.e. to understand the data in the context of an application domain (see > Research statement).
- This needs an international concerted effort (see > Expert network) and education of a new kind of graduates (see > Teaching statement). Cross-domain approaches foster serendipity, cross-fertilize methodologies and insights, and ultimately transfer ideas into Business for the benefit of humans (see > Conference CD-MAKE).