Signal Processing and Speech Communication Laboratory
hometheses & projects › Modelling of gas sensor responses using deep models

Modelling of gas sensor responses using deep models

Status
Open
Type
Master Thesis
Announcement date
09 Mar 2015
Mentors
Research Areas

Materials used in gas sensors typically feature a resistance sensitive to atmospheric constituents and respective concentration as a nonlinear function of temperature and humidity. Time resolved resistance data of temperature transitions provides rich yet compact information for sensor characterisation. In this work, deep models shall be explored to model sensor material responses. Real-world data sets provide an excellent opportunity to link current materials research with latest machine learning technology.