Combining Machine Learning and Formal Techniques for Small Data Applications - A Framework to Explore New Structural Materials

Rolf Drechsler, Sebastian Huhn, Christina Plump

In: Euromicro Conference on Digital System Design (DSD). Euromicro Conference on Digital System Design (DSD-2020) August 26-28 Portoro¸ Slovenia 2020.


The massive increase in computation power leads to a renaissance of supervised learning techniques, which were published decades ago but have so far been confined to theory. These techniques form the increasingly important field of Ma-chine Learning(ML), which contributes to a large variety of research concerning industrial, automotive but also consumer applications strongly influencing our daily life. Commonly, the learning techniques require a set of labeled data, which involves are source-intensive generation, to conduct the training. Depending on the dimensionality of the data and the required precision as needed by the application, the amount of training data varies. Incase of insufficient training data, the prediction is of low-quality or not even possible at all, restricting the applicability of ML. This work proposes a combination of formal techniques and ML to implement a framework that allows coping with high-dimensional, training data while retaining a high predictionquality. The efficacy of this method is exemplarily demonstratedon the basis of an interdisciplinary material science researchproblem concerning the development of new structural materials,though it can be adapted to further applications.

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence