Graduate School and Research Center in Digital Sciences

Using deep learning to replace domain knowledge

Lubben, Christian; Pah, Marc-Oliver; Khan, Mohammad Irfan

ISCC 2020, 25th IEEE Symposium on Computers and Communications, 7-10 July, 2020, Rennes, France

Complex problems like the prediction of future behavior of a system are usually solved by using domain knowledge. This knowledge comes with a certain expense which can be monetary costs or efforts to generate it. We want to decrease this cost while using state of the art machine learning and prediction methods. Our aim is to replace the domain knowledge and create a black-box solution that offers automatic reasoning and accurate predictions. Our guiding example is packet scheduling optimization in Vehicle to Vehicle (V2V) communication. Within the evaluation, we compare the prediction quality of a labourintense whitebox approach with the presented fully-automated blackbox approach. To ease the measurement process we propose a framework design which allows easy exchange of predictors. The results show the successful design of our framework as well as superior accuracy of the black box approach.

Document Bibtex

Title:Using deep learning to replace domain knowledge
Keywords:V2V, V2X, network traffic prediction, deep learning, ANN
Department:Communication systems
Eurecom ref:6322
Copyright: © 2020 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Bibtex: @inproceedings{EURECOM+6322, year = {2020}, title = {{U}sing deep learning to replace domain knowledge}, author = {{L}ubben, {C}hristian and {P}ah, {M}arc-{O}liver and {K}han, {M}ohammad {I}rfan}, booktitle = {{ISCC} 2020, 25th {IEEE} {S}ymposium on {C}omputers and {C}ommunications, 7-10 {J}uly, 2020, {R}ennes, {F}rance}, address = {{R}ennes, {FRANCE}}, month = {07}, url = {} }
See also: