This thesis presents a container based execution stack for neural networks (ConbexNN) using the Kubernetes container orchestration and a Java based microservice architecture, which is exposed to users and other systems via RESTful webservices. The whole workflow including importing, training and evaluating a neural network model, becomes possible by using this service oriented approach. This work is influenced by N2Sky, a framework for the exchange of neural network specific knowledge and aims to support ViNNSL, the Vienna Neural Network Specification Language. The execution stack runs on many common cloud platforms. Furthermore it is scalable and each component is extensible and interchangeable.
Masterprüfung mit Defensio, Nussbaum Benjamin
18.10.2018 13:00 - 14:30
Organiser:
SPL 5
Location:
Besprechungsraum 4.34
Währinger Straße 29
1090 Wien
1090 Wien