Uncertainty Quantification for Deep Context-Aware Mobile Activity Recognition and Unknown Context Discovery

Zepeng Huo, Arash PakBin, Xiaohan Chen, Nathan Hurley, Ye Yuan, Xiaoning Qian, Zhangyang Wang, Shuai Huang, Bobak Mortazavi
Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, PMLR 108:3894-3904, 2020.

Abstract

Activity recognition in wearable computing faces two key challenges: i) activity characteristics may be context-dependent and change under different contexts or situations; ii) unknown contexts and activities may occur from time to time, requiring flexibility and adaptability of the algorithm. We develop a context-aware mixture of deep models termed the $\alpha$-$\beta$ network coupled with uncertainty quantification (UQ) based upon maximum entropy to enhance human activity recognition performance. We improve accuracy and F score by 10% by identifying high-level contexts in a data-driven way to guide model development. In order to ensure training stability, we have used a clustering-based pre-training in both public and in-house datasets, demonstrating improved accuracy through unknown context discovery.

Cite this Paper

BibTeX
@InProceedings{pmlr-v108-huo20a, title = {Uncertainty Quantification for Deep Context-Aware Mobile Activity Recognition and Unknown Context Discovery}, author = {Huo, Zepeng and PakBin, Arash and Chen, Xiaohan and Hurley, Nathan and Yuan, Ye and Qian, Xiaoning and Wang, Zhangyang and Huang, Shuai and Mortazavi, Bobak}, booktitle = {Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics}, pages = {3894--3904}, year = {2020}, editor = {Chiappa, Silvia and Calandra, Roberto}, volume = {108}, series = {Proceedings of Machine Learning Research}, month = {26--28 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v108/huo20a/huo20a.pdf}, url = {https://proceedings.mlr.press/v108/huo20a.html}, abstract = {Activity recognition in wearable computing faces two key challenges: i) activity characteristics may be context-dependent and change under different contexts or situations; ii) unknown contexts and activities may occur from time to time, requiring flexibility and adaptability of the algorithm. We develop a context-aware mixture of deep models termed the $\alpha$-$\beta$ network coupled with uncertainty quantification (UQ) based upon maximum entropy to enhance human activity recognition performance. We improve accuracy and F score by 10% by identifying high-level contexts in a data-driven way to guide model development. In order to ensure training stability, we have used a clustering-based pre-training in both public and in-house datasets, demonstrating improved accuracy through unknown context discovery.} }
Endnote
%0 Conference Paper %T Uncertainty Quantification for Deep Context-Aware Mobile Activity Recognition and Unknown Context Discovery %A Zepeng Huo %A Arash PakBin %A Xiaohan Chen %A Nathan Hurley %A Ye Yuan %A Xiaoning Qian %A Zhangyang Wang %A Shuai Huang %A Bobak Mortazavi %B Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2020 %E Silvia Chiappa %E Roberto Calandra %F pmlr-v108-huo20a %I PMLR %P 3894--3904 %U https://proceedings.mlr.press/v108/huo20a.html %V 108 %X Activity recognition in wearable computing faces two key challenges: i) activity characteristics may be context-dependent and change under different contexts or situations; ii) unknown contexts and activities may occur from time to time, requiring flexibility and adaptability of the algorithm. We develop a context-aware mixture of deep models termed the $\alpha$-$\beta$ network coupled with uncertainty quantification (UQ) based upon maximum entropy to enhance human activity recognition performance. We improve accuracy and F score by 10% by identifying high-level contexts in a data-driven way to guide model development. In order to ensure training stability, we have used a clustering-based pre-training in both public and in-house datasets, demonstrating improved accuracy through unknown context discovery.
APA
Huo, Z., PakBin, A., Chen, X., Hurley, N., Yuan, Y., Qian, X., Wang, Z., Huang, S. & Mortazavi, B.. (2020). Uncertainty Quantification for Deep Context-Aware Mobile Activity Recognition and Unknown Context Discovery. Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 108:3894-3904 Available from https://proceedings.mlr.press/v108/huo20a.html.

Related Material