Successful IML4E Plenary in Helsinki
Over the past three days, the first physical IML4E plenary meeting has taken place in Helsinki. After almost a year of project work under corona conditions, our three-day meeting allowed us to see how productive meetings can be when people actually sit face-to-face. We were able to enjoy three full days of exciting discussions about our case studies, their requirements and the technical solutions planned in the project, as well as the related research. With 20 people on-site and another 10 people online, we discussed issues of data preparation, data refinement, and ML in industrial use.
For example, the Software AG presented its method of meta-feature learning, Fraunhofer FOKUS explained its approach for continuous quality assurance of ML, the University of Helsinki reported, among others, on a method for error detection in ML models based on the analysis of node co-activations, and the company SpiceTech presented its tool VALICY, which can be used to identify critical decision boundaries in ML models.
The project partners discussed in a concentrated atmosphere in which form the individual partner technologies can be used in order to meet the requirements of the case study partners. For more information about the project please follow us on Twitter, on LinkedIn or just visit our website.