The paper, “Network Scheduling and Compute Resource Aware Task Placement in Datacenters“, is the journal version of an earlier conference paper with my intern Ali Munir at CoNEXT’16. It talks about the importance of taking into account how the network schedules flows in placing tasks for data-intensive computation in datacenters.
Hanlin’s preliminary result in exploring the data privacy aspect of coreset-based learning, in comparison with federated learning, is accepted for a long presentation (11 out of 46) at the International Workshop on Federated Learning for User Privacy and Data Confidentiality in Conjunction with ICML 2020 (FL-ICML’20). Congratulations, Hanlin!
Quinn’s paper on “Misreporting Attacks in Software-Defined Networking” is accepted to International Conference on Security and Privacy in Communication Networks (SecureComm 2020). Congratulations, Quinn! The work is supported by a NSF:SaTC grant for which Ting He is the PI. It talks about how a compromised switch can stealthily attract a greater portion of traffic than […]
Our work on supporting diverse machine learning models by collecting a robust coreset was featured in the article “Machine learning algorithms promise better situational awareness“, published by U.S. Army CCDC Army Research Laboratory Public Affairs. The work is part of Hanlin’s thesis research, originally published in Globecom’19 and JSAC’20. We thank our government partner Dr. […]
Sina’s paper “Solving the Divergence Problem in AC-QSS Cascading Failure Model by Introducing the Effect of a Realistic UVLS Scheme” develops a AC-QSS based model which better captures the process of cascading failure in a smart grid by incorporating the effect of a built-in failure mitigation scheme called Under Voltage Load Shedding (UVLS). This is […]
Chronically, this is Hanlin’s second piece of work done during his internship at IBM last summer. The work tries, for the first time, to combine coreset and quantization to jointly reduce two dimensions of the data cube: the number of data points, and the number of bits in representing each feature of each data point. […]
His thesis is on “Flow Table Security” where he formulated a novel inference problem called Adversarial Cache Inference problem about probing, inferring, and attacking a flow table (modeled as a cache) from one or a set of compromised host(s). The work has been accepted to INFOCOM’20 and the long version is under submission to IEEE/ACM […]
Hanlin’s very first full-sized work, Robust Coreset Construction for Distributed Machine Learning, finally got accepted into JSAC Special Issue on Advances in Artificial Intelligence and Machine Learning for Networking. Congratulations, Hanlin! I will share the camera-ready version later, but the main idea is already captured in this arXiv version. It talks about how to use […]
Hanlin’s paper on “Communication-efficient k-Means for Edge-based Machine Learning” is accepted to ICDCS’20. It continues our line of investigation on robust and efficient distributed machine learning via the approach of “sharing data summaries”. Congratulations, Hanlin!
Yilei’s paper on “Looking Glass of NFV: Inferring the Structure and State of NFV Network from External Observations” is accepted to IEEE/ACM Transactions on Networking. Congratulations, Yilei! This is the full version of her INFOCOM’19 paper under the same title.