Machine Learning for Networking/Robotics/Healthcare/Manufacturing Applications:
Reinforcement Learning for Network Resource Allocation:
Reinforcement Learning has been increasingly used for efficient resource allocation. However, there are multiple challenges in networking applications that make the reinforcement learning tools not applicable. Some examples include multiple agents where the fairness objective causes non-linearity, scaling to large networks leads to efficient solutions splitting the network, constraints in decision making, etc. In this theme, we propose novel algorithms to take these aspects into account for efficient algorithms for traffic engineering, power control, and caching decisions.
- Nan Geng, Tian Lan, Vaneet Aggarwal, Yuan Yang, and Mingwei Xu, "A Multi-agent Reinforcement Learning Perspective on Distributed Traffic Engineering," in Proc. IEEE International Conference on Network Protocols (ICNP), Oct 2020. (16.8% acceptance rate, 31/184).
- Yimeng Wang, Yongbo Li, Tian Lan, and Vaneet Aggarwal, "DeepChunk: Deep Q-Learning for Chunk-based Caching in Data Processing Networks," IEEE Transactions on Cognitive Communications and Networking, Special Issue on Deep Reinforcement Learning for Future Wireless Communication Networks, vol. 5, no. 4, pp. 1034-1045, Dec. 2019.
- Ramkumar Raghu, Pratheek Upadhyaya, Mahadesh Panju, Vaneet Aggarwal, and Vinod Sharma, "Deep Reinforcement Learning Based Power control for Wireless Multicast Systems," in Proc. Allerton, Oct 2019.
- Yimeng Wang, Yongbo Li, Vaneet Aggarwal, and Tian Lan, "Deep Q-Learning for Chunk-based Caching in Data Processing Networks," in Proc. Allerton, Oct 2019.
Teleoperated Robotic Surgery via Transfer Learning:
In austere environments, teleoperated surgical robots could save the lives of critically injured patients if they can perform complex surgical maneuvers under limited communication bandwidth. The bandwidth requirement is reduced by transferring atomic surgical actions (referred to as “surgemes") instead of the low-level kinematic information. While such a policy reduces the bandwidth requirement, it requires accurate recognition of the surgemes. In this work, we demonstrate that transfer learning across surgical tasks can boost the performance of surgeme recognition.
- Mridul Agarwal and Vaneet Aggarwal, "Blind Decision Making: Reinforcement Learning with Delayed Observations," in Proc. ICAPS, Aug 2021.
- Mridul Agarwal, Glebys Gonzalez, Mythra V. Balakuntala, Md Masudur Rahman, Vaneet Aggarwal, Richard M. Voyles, Yexiang Xue, and Juan Wachs, "Dexterous Skill Transfer between Surgical Procedures for Teleoperated Robotic Surgery," in Proc. 30th IEEE International Conference on Robot & Human Interactive Communication (RO-MAN), Aug 2021.
- Md Masudur Rahman, Mythra Varun Balakuntala Srinivasa Mur, Mridul Agarwal, Upinder Kaur, Vishnunandan Lakshmi Venkatesh, Glebys Gonzalez, Natalia Sanchez Tamayo, Yexiang Xue, Richard Voyles, Vaneet Aggarwal, and Juan Wachs, "SARTRES: A Semi-Autonomous Robot TeleopeRation Environment for Surgery," AE-CAI 2020 Special Issue of the Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization Journal (TCIV), Nov 2020, DOI: 10.1080/21681163.2020.1834878.
- Glebys Gonzalez, Mridul Agarwal, Mythra Varun Balakuntala Srinivasa Murthy, Md Masudur Rahman, Upinder Kaur, Juan Wachs, Richard Voyles, Vaneet Aggarwal, and Yexiang Xu, "DESERTS:Delay-Tolerant Semi-Autonomous Robot Teleoperation for Surgery," in Proc. IEEE International Conference on Robotics and Automation (ICRA), May-Jun 2021.
- MMd Masudur Rahman, Natalia Sanchez-Tamayo, Glebys Gonzalez, Mridul Agarwal, Vaneet Aggarwal, Richard Voyles, Yexiang Xue, Juan Wachs, "Transferring Dexterous Surgical Skill Knowledge between Robots for Semi-autonomous Teleoperation," in Proc. IEEE International Conference on Robot and Human Interactive Communication (Ro-Man), Oct 2019.
Machine Learning for Manufacturing:
Today's manufacturing starts with CAD models. The CAD model carries all the geometrical information of the workpiece (e.g., surface position, size, fluctuations etc.) and influences the entire planning for manufacturing processes. Due to the complexity of an engineering CAD model, the manufacturing industry usually decomposes the CAD model into several geometrical features - called the machining features. The machining features serve certain functions of the final product and also require typical machining processes. This CAD model decomposition and feature-recognition processes together are called machining feature identification, which is the focus of our work, where efficient data representation techniques and machine learning approaches are used.
- Xingyu Fu, Fengfeng Zhou, Dheeraj Peddireddy, Zhengyang Kang, Martin Byung-Guk Jun, and Vaneet Aggarwal, "An FEA surrogate model with Boundary Oriented Graph Embedding approach," arXiv, Aug 2021
- Xingyu Fu, Dheeraj Peddireddy, Vaneet Aggarwal, and Martin Byung-Guk Jun, "Improved Dexel Representation: A 3D CNN Geometry Descriptor for Manufacturing CAD,," Accepted to IEEE Transactions on Industrial Informatics, Dec 2021.
- Dheeraj Peddireddy, Xingyu Fu, Anirudh Shankar, Haobo Wang, Byung Gun Joung, Vaneet Aggarwal, John W. Sutherland, and Martin Byung-Guk Jun, "Identifying Manufacturability and Machining Processes using Deep 3D Convolutional Networks," Journal of Manufacturing Processes, vol. 64, pp. 1336-1348, Apr 2021.
- Dheeraj Peddireddy, Xingyu Fu, Haobo Wang, Byung Gun Joung, Vaneet Aggarwal, John W. Sutherland, and Martin Byung-Guk Jun, "Deep Learning Based Approach for Identifying Conventional Machining Processes from CAD Data ," in Proc. NAMRC, Jun 2020.
Machine Learning for Healthcare:
Using efficient learning based techniques are essential for predicting risk in individuals. In many cases, passive face videos can be used to predict the health risks. We have used learning based techniques for early prediction of Sepsis, prediction of lifting load risk, force exertions, and health monitoring.
- Guoyang Zhou, Vaneet Aggarwal, Ming Yun, and Denny Yu, "A Computer Vision Approach for Estimating Lifting Load Contributors to Injury Risk," Accepted to IEEE Transactions on Human-Machine Systems, Jan 2022.
- Guoyang Zhou, Vaneet Aggarwal, Ming Yun, and Denny Yu, "Video-Based AI Decision Support System for Lifting Risk Assessment,," in Proc. IEEE SMC, Oct. 2021.
- Naimahmed Nesaragi, Shivnarayan Patidar, and Vaneet Aggarwal, "Tensor Learning of Pointwise Mutual Information from EHR Data for Early Prediction of Sepsis," Computers in Biology and Medicine, Volume 134, July 2021, 104430.
- Hamed Asadi, Guoyang Zhou, Jae Joong Lee, Vaneet Aggarwal, and Denny Yu, "A Computer Vision Algorithm to Identify High Force Exertions from Facial Expressions," Ergonomics, Apr 2020.
- Mayank Gupta, Lingjun Chen, Denny Yu, and Vaneet Aggarwal, "A Supervised Learning Approach for Robust Health Monitoring using Face Videos," in Proc. 2nd ACM Workshop on Device Free Human Sensing (DFHS, ACM Buildsys Workshop), Nov. 2020
Applications of Exploiting Data Structure in Matrix/Tensor for Better Sampling, Fingerprinting, and Estimation:
Exploiting structure in the data across multiple dimensions can lead to efficient techniques for efficient sampling and fingerprinting. The figure alongside shows different network service atributes dependence on time, spatial locations and data service quality measurements thus motivating tensor structure. Further, the map shows the received signal strength in indoor has spatial dependence and the signals from different APs are correlated and thus adaptive sampling can reduce fingerprints needed for localization.
- X. Liu, S. Aeron, V. Aggarwal, and X. Wang, "Low-tubal-rank Tensor Completion using Alternating Minimization," IEEE Transactions on Information Theory, vol. 66, no. 3, pp. 1714-1737, March 2020.
- W. Wang, Y. Sun, B. Eriksson, W. Wang, and V. Aggarwal, "Wide Compression: Tensor Ring Nets," in Proc. CVPR, Jun 2018
- W.Wang, V. Aggarwal, and S. Aeron, "Efficient Low Rank Tensor Ring Completion," in Proc. ICCV, Oct 2017.
- W.Wang, V. Aggarwal, and S. Aeron, "Unsupervised Clustering Under The Union of Polyhedral Cones (UOPC) Model," Pattern Recognition Letters, vol. 100, pp. 104-109, Dec 2017.
- V. Aggarwal, A. A. Mahimkar, H. Ma, Z. Zhang, S. Aeron, and W. Willinger, "Inferring Smartphone Service Quality using Tensor Methods," in Proc. 12th International Conference on Network and Service Management Oct-Nov, 2016.
- X. Liu, S. Aeron, V. Aggarwal, X. Wang, and M. Wu, "Adaptive Sampling of RF fingerprints for Fine-grained Indoor Localization," IEEE Transactions on Mobile Computing, vol. 15, no. 10, pp. 2411-2423, Oct. 2016.
- X. Liu, S. Aeron, V. Aggarwal, and X. Wang, "Low-tubal-rank Tensor Completion using Alternating Minimization," in Proc. SPIE Defense + Security, 2016.
- X. Liu, S. Aeron, V. Aggarwal, X. Wang, and M. Wu, "Tensor completion via adaptive sampling of tensor fibers: An application to efficient RF fingerprinting," in Proc. IEEE ICASSP, Mar 2016.
Exploiting Network Structure for Better Scheduling Strategies:
Learning the network characteristics is crucial for coming up with next generation network designs. Getting the measurements of the received power from a tower leads to path loss which helps in design of efficient codes. Fingerprinting roads can help find efficient scheduling/handoff decisions for a service provider. Learning attributes at the network helps learn interaction between different layers to achieve better designs. The distribution of interference can lead to better interference management techniques. This understanding will be useful to achieve new and better system designs. The figure alongside depicts a mobile user trajectory along a road through 3 cellular sectors showing interactions with the cellular network and the measured values of channel quality (Ec/Io) during 3 different drives. This structure has been used to novel scheduling strategies with prediction.
- R. Margolies, A. Sridharan, V. Aggarwal, R. Jana, N. K. Shankaranayanan, V. Vaishampayan, and G. Zussman, "Exploiting Mobility in Proportional Fair Cellular Scheduling: Measurements and Algorithms," IEEE/ACM Transactions on Networking, vol. 24, no. 1, pp. 355-367, Feb. 2016.
- R. Margolies, A. Sridharan, V. Aggarwal, R. Jana, N. K. Shankaranayanan, V. Vaishampayan, and G. Zussman, "Exploiting Mobility in Proportional Fair Cellular Scheduling: Measurements and Algorithms," in Proc. Infocomm 2014.
- V. Aggarwal, R. Jana, J. Pang, K. K. Ramakrishnan and N. K. Shankaranarayanan, "Characterizing Fairness for 3G Wireless Networks," in Proc. LANMAN, Oct. 2011.
- V. Aggarwal and S. Krishnan, "Achieving Approximate Soft Clustering in Data Streams," 2011. Implementation
Modeling Application QoE from Passive Network Measurements:
Video and web applications are the most prominent applications in today's world. It is important to understand how radio network characteristics (such as signal strength, handovers, load, etc.) influence users' Quality-of-Experience (QoE). Understanding the relationship between QoE and network characteristics is a pre-requisite for cellular network operators to detect when and where degraded network conditions actually impact QoE. Unfortunately, cellular network operators do not have access to detailed server-side or client-side logs to directly measure QoE metrics, such as abandonment rate, stalls, and video/session length. We devise a machine-learning-based mechanism to infer web QoE metrics from network traces accurately. The image alongside depicts partial web downloads as a function of network attributes.
- A. Balachandran, V. Aggarwal, E. Halepovic, J. Pang, S. Seshan, S. Venkataraman, and H. Yan "Modeling Web Quality-of-Experience on Cellular Networks," in Proc. Mobicom 2014
- V. Aggarwal, E. Halepovic, J. Pang, S. Venkataraman, and H. Yan "Prometheus: Toward Quality-of-Experience Estimation for Mobile Apps from Passive Network Measurements," in Proc. Hotmobile 2014
- V. Ramaswami, K. Jain, R. Jana, and V. Aggarwal, "Modeling Heavy-tails in Traffic Sources for Network Performance Evaluation," in Proc. International Conference on Computational Intelligence, Cyber Security and Computational Models, Dec. 2013.