论文信息 - 4.3 - Application-Specific Partially Automated Design of Multi-Sensor Intelligent Lab-on-Spoon System

4.3 - Application-Specific Partially Automated Design of Multi-Sensor Intelligent Lab-on-Spoon System

In this work, we pick up the ISE research on automation of intelligent integrated systems design and apply and adapt the concept, methodology, and tool implementation. In particular, a new platform, based on the proven machinelearning tool ORANGE will make the work multi-platform and open-access. We demonstrate the current status by the application to our Lab-on-Spoon (LoS) multi-sensory system and corresponding data. The creation of a an efficient Support-Vector-Machine (SVM)-based hierarchical classification architecture and options of automated feature selection (AFS) are investigated in the experiments. An improved in two cases to 100% classification accuracy with regard to the previous flat approach was achieved.

A. König | K. Thongpull | A. König | K. Thongpull

[1] Li Li,et al. Can Impedance Spectroscopy serve in an embedded multi-sensor system to improve driver drowsiness detection? , 2015 .

[2] Gunnar Rätsch,et al. An introduction to kernel-based learning algorithms , 2001, IEEE Trans. Neural Networks.

[3] Martin Mozina,et al. Orange: data mining toolbox in python , 2013, J. Mach. Learn. Res..

[4] John Platt,et al. Probabilistic Outputs for Support vector Machines and Comparisons to Regularized Likelihood Methods , 1999 .

[5] Hiroshi Motoda,et al. Feature Selection for Knowledge Discovery and Data Mining , 1998, The Springer International Series in Engineering and Computer Science.

[6] Chih-Jen Lin,et al. LIBSVM: A library for support vector machines , 2011, TIST.

[7] Ron Kohavi,et al. Feature Selection for Knowledge Discovery and Data Mining , 1998 .

[8] Andreas König,et al. Methodology, Algorithms, and Emerging Tool for Automated Design of Intelligent Integrated Multi-Sensor Systems , 2009, Algorithms.