top of page

Workshop on Streaming Systems and Realtime Machine Learning (STREAM-ML)



As a successor to the STREAM 2015 & 2016 workshops, STREAM-ML aims to advance decision-making and control over complex systems by applying machine learning techniques to streams of real-time data.


Applications associated with streaming data and related real-time machine learning, steering and control are of growing interest and importance. The analysis of data streaming from on-line instruments, large scale simulations, and Internet of Things (IoT), an enormous amount of distributed sensors now enables near real-time steering and control of complex systems such as scientific experiments, transportation systems, and urban environments. STREAM-ML has the potential to provide more timely access to information and raise the quality and pace of decision making and, consequently, performance. There has been an explosion of new research and technologies for stream analytics arising from the academic and private sectors. We need to research steering scenarios involving active learning. Steering, which is inevitably real-time, might include realigning experimental sensors, control of autonomous vehicle or changing progress of simulations.

Advances in ML techniques, such as Deep Learning and neural networks, are the main enablers of knowledge work automation. Natural user interfaces, such as speech and image recognition are also highly benefiting from ML technologies. Compared to the economic impact of machine learning, next frontier is knowledge work automation that asserts the more attention toward the extraction of value out of data.

Program Schedule

Tuesday, December 10, 2019
Room: Echo Park
Westin Bonaventure Hotel & Suites Located at 404 South Figueroa Street, Los Angeles, CA

Workshop Topics
  • Real-Time ML Applications for social media, sports, healthcare, financial transactions

  • Real-Time ML Applications for IoT, Cyberphysical Systems, Satellite and airborne monitors

  • Real-Time ML Applications for instruments like the LHC, Sequencers, Data Assimilation

  • Real-Time ML Applications for Astronomy, Light Sources, climate and Agriculture

  • Streaming ML and Analysis of Simulation Results

  • Data fusion and reduction for IoT devices

  • Distributed machine learning in a cloud environment

  • Streaming batch and online learning algorithms (CNN, RNN, LSTM, GAN, Neuromorphic)

  • Real-Time Training and inference

  • Distributed Machine Learning

  • Approximation algorithms for IoT devices

  • Programming and Runtime Model

  • Streaming Software Systems and Algorithm Library

  • Streaming Infrastructure on major commercial clouds (CPU, GPU, FPGA, TPU)

  • Streaming ML on Edge Devices

  • Streaming ML on IoT Devices

  • Steering and Human in the Loop

  • Real-Time Steering and Control

  • Metrics of performance and benchmarks

  • Security for Real-Time ML

Call for Papers

Members of the community are invited to submit a paper (6 pages for a full paper or 2 pages for an extended abstract) in areas of relevance to STREAM-ML's scope and objectives. The paper format is a two-column IEEE Conference Style. Papers must be electronically submitted using the submission system accessible on CyberChair.

All papers accepted will be included in the IEEE Big Data Conference Proceedings published by the IEEE Computer Society Press. At least one author of each accepted paper must register for the conference and present the paper at the workshop for the paper to be included in the conference proceedings. Details on the registration will be posted on the main conference's page.

bottom of page