Veröffentlichung: 2019 Juli
Art der Veröffentlichung: https://arxiv.org/abs/1907.11569
Research on neural networks has gained significant momentum over the past few years. Because training is a resource-intensive process and training data cannot always be made available to everyone, there has been a trend to reuse pre-trained neural networks. As such, neural networks themselves have become research data. In this paper, we first present the FAIRnets Ontology, an ontology to make existing neural network models findable, accessible, interoperable, and reusable (see the FAIR principles for scientific data management and stewardship). Our ontology allows us to model neural networks on a meta-level in a structured way, including the representation of all network layers and their characteristics. Secondly, we have modeled over 18,400 neural networks based on this ontology, which we provide to the public on GitHub as a knowledge graph called FAIRnets for research purposes. Among other use cases, FAIRnets can be used by researchers and practitioners (e.g., data scientists) for finding suitable pre-trained neural networks given specific requirements by the users.
Weitere Informationen unter: Link