Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Influential factors on technology sourcing for the use of deep learning

: Masior, Jonathan; Kara, Gözde; Schimpf, Sven

Volltext urn:nbn:de:0011-n-4700196 (679 KByte PDF)
MD5 Fingerprint: d0c334b854051f1bb68df2da23c9ad04
Erstellt am: 19.10.2017

University of Cambridge, Institute for Manufacturing:
Science, markets & society. Crossing boundaries, creating momentum : R&D Management Conference 2017; Leuven, 1 - 5 July 2017
Leuven, 2017
11 S.
R&D Management Conference <2017, Leuven>
Konferenzbeitrag, Elektronische Publikation
Fraunhofer IAO ()

Various companies found an effective way to outsource their development efforts in open source communities. The availability and easy accessibility of resources such as high computational capability and distributed processing, open source software, and sophisticated libraries, enable competitions within open source communities to develop the best predictions with deep learning algorithms. To train and to develop learning models this ecosystem requires high quality, consistent, structured data. The research on technology sourcing decisions is a well-documented realm, which automated software still did not penetrate. Influential factors on the sourcing decision of product technologies derived from an in-depth literature analysis. The essential requirements for models of sourcing decisions and intelligent decision support algorithms are juxtaposed. The purpose was to respond to current trends and develop a logical structure for datasets significant as an input to deep learning software. The resulting models are applicable on various complex decisions and analyses in strategic technology management by communities, researchers, and companies, which develop and train learning models based on their own, individual preferences.