Data-Guided Approach for Learning and Improving User Experience in Computer Networks

Yanan Bao, Xin Liu, Amit Pande
Asian Conference on Machine Learning, PMLR 45:127-142, 2016.

Abstract

Machine learning algorithms have been traditionally used to understand user behavior or system performance. In computer networks, with a subset of input features as controllable network parameters, we envision developing a data-driven network resource allocation framework that can optimize user experience. In particular, we explore how to leverage a classifier learned from training instances to optimally guide network resource allocation to improve the overall performance on test instances. Based on logistic regression, we propose an optimal resource allocation algorithm, as well as heuristics with low-complexity. We evaluate the performance of the proposed algorithms using a synthetic Gaussian dataset, a real world dataset on video streaming over throttled networks, and a tier-one cellular operator’s customer complaint traces. The evaluation demonstrates the effectiveness of the proposed algorithms; e.g., the optimal algorithm can have a 400% improvement compared with the baseline.

Cite this Paper

BibTeX
@InProceedings{pmlr-v45-Bao15, title = {Data-Guided Approach for Learning and Improving User Experience in Computer Networks}, author = {Bao, Yanan and Liu, Xin and Pande, Amit}, booktitle = {Asian Conference on Machine Learning}, pages = {127--142}, year = {2016}, editor = {Holmes, Geoffrey and Liu, Tie-Yan}, volume = {45}, series = {Proceedings of Machine Learning Research}, address = {Hong Kong}, month = {20--22 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v45/Bao15.pdf}, url = {https://proceedings.mlr.press/v45/Bao15.html}, abstract = {Machine learning algorithms have been traditionally used to understand user behavior or system performance. In computer networks, with a subset of input features as controllable network parameters, we envision developing a data-driven network resource allocation framework that can optimize user experience. In particular, we explore how to leverage a classifier learned from training instances to optimally guide network resource allocation to improve the overall performance on test instances. Based on logistic regression, we propose an optimal resource allocation algorithm, as well as heuristics with low-complexity. We evaluate the performance of the proposed algorithms using a synthetic Gaussian dataset, a real world dataset on video streaming over throttled networks, and a tier-one cellular operator’s customer complaint traces. The evaluation demonstrates the effectiveness of the proposed algorithms; e.g., the optimal algorithm can have a 400% improvement compared with the baseline. } }
Endnote
%0 Conference Paper %T Data-Guided Approach for Learning and Improving User Experience in Computer Networks %A Yanan Bao %A Xin Liu %A Amit Pande %B Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Geoffrey Holmes %E Tie-Yan Liu %F pmlr-v45-Bao15 %I PMLR %P 127--142 %U https://proceedings.mlr.press/v45/Bao15.html %V 45 %X Machine learning algorithms have been traditionally used to understand user behavior or system performance. In computer networks, with a subset of input features as controllable network parameters, we envision developing a data-driven network resource allocation framework that can optimize user experience. In particular, we explore how to leverage a classifier learned from training instances to optimally guide network resource allocation to improve the overall performance on test instances. Based on logistic regression, we propose an optimal resource allocation algorithm, as well as heuristics with low-complexity. We evaluate the performance of the proposed algorithms using a synthetic Gaussian dataset, a real world dataset on video streaming over throttled networks, and a tier-one cellular operator’s customer complaint traces. The evaluation demonstrates the effectiveness of the proposed algorithms; e.g., the optimal algorithm can have a 400% improvement compared with the baseline.
RIS
TY - CPAPER TI - Data-Guided Approach for Learning and Improving User Experience in Computer Networks AU - Yanan Bao AU - Xin Liu AU - Amit Pande BT - Asian Conference on Machine Learning DA - 2016/02/25 ED - Geoffrey Holmes ED - Tie-Yan Liu ID - pmlr-v45-Bao15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 45 SP - 127 EP - 142 L1 - http://proceedings.mlr.press/v45/Bao15.pdf UR - https://proceedings.mlr.press/v45/Bao15.html AB - Machine learning algorithms have been traditionally used to understand user behavior or system performance. In computer networks, with a subset of input features as controllable network parameters, we envision developing a data-driven network resource allocation framework that can optimize user experience. In particular, we explore how to leverage a classifier learned from training instances to optimally guide network resource allocation to improve the overall performance on test instances. Based on logistic regression, we propose an optimal resource allocation algorithm, as well as heuristics with low-complexity. We evaluate the performance of the proposed algorithms using a synthetic Gaussian dataset, a real world dataset on video streaming over throttled networks, and a tier-one cellular operator’s customer complaint traces. The evaluation demonstrates the effectiveness of the proposed algorithms; e.g., the optimal algorithm can have a 400% improvement compared with the baseline. ER -
APA
Bao, Y., Liu, X. & Pande, A.. (2016). Data-Guided Approach for Learning and Improving User Experience in Computer Networks. Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 45:127-142 Available from https://proceedings.mlr.press/v45/Bao15.html.

Related Material