Wrinkles in Time: Detecting Internet-wide Events via NTP

Syamkumar, Meenakshi; Mani, Sathiya Kumaran; Durairajan, Ramakrishnan; Barford, Paul; Sommers, Joel

doi:10.23919/ifipnetworking.2018.8696566

Cited by 5 publications

(1 citation statement)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These techniques construct predictive models by learning from a large number of training examples (i.e., labeled data) but as discussed earlier, the networking domain lacks in general access to the necessary training data. Similarly, several efforts use unsupervised learning [43] to e.g., detect anomalies in BGP [25], perform network traffic prediction and diagnosis [14,[26][27][28], or carry out event detection [42]. However, the problem with unsupervised learning is that not all types of clustering techniques are suitable for identifying events of interest in networking data.…”

Section: Prior Efforts and Their Limitationsmentioning

confidence: 99%

Challenges in Using ML for Networking Research

Lavinia¹,

Durairajan²,

Rejaie³

et al. 2020

Proceedings of the Workshop on Network Meets AI &Amp; ML

Self Cite

View full text Add to dashboard Cite

Leveraging innovations in Machine Learning (ML) research is of great current interest to researchers across the sciences, including networking research. However, using ML for networking poses challenging new problems that have been responsible for slowing the pace of innovation and the adoption of ML in the networking domain. Among the main problems are a well-known lack of data in general and representative data in particular, an overall inability to label data at scale, unknown data quality due to differences in data collection strategies, and data privacy issues that are unique to network data. Motivated by these challenges, we describe the design of Emerge 1 , a novel framework to support efforts to dEmocratize the use of ML for nEtwoRkinG rEsearch. In particular, Emerge focuses on the problem of providing a low-cost, scalable, and highquality methodology for labeling networking data. To illustrate the benefits of Emerge, we use publicly available network measurement datasets from Caida's Ark project and create and evaluate data labels for them in a programmable fashion.

show abstract