IEEE Signal Processing Society Summer School on Signal Processing and Machine Learning for Big Data at the University of Pittsburgh

Ervin Sejdić and Murat Akcakaya

Big data and machine learning are buzz terms that we frequently hear within the scientific and industrial societies and read about them in many scientific publications over the last several years. Nowadays, recent technological advancements have made recording and streaming large amount of data a reality, but processing, storing and communicating such Big Data have become the main challenges that many industries have to face. In recent years, companies are seeking to hire undergraduate and graduate students that have sufficient skills covering big data and/or machine learning. Driven by this industrial need, we decided to organize a summer school at the University of Pittsburgh that brings together the academic and industrial leaders to discuss the Big Data needs and solutions.

With support from the society leadership, we began a daunting task of organizing the event in January 2016. Knowing that the summer school should be held closely to the end of academic semester at most American and Canadian institutions, we decided on May 17th-19th, 2016 as suitable time for the summer school, not too far into the summer months to interfere with other conferences and everyone’s vacations plans. While we had settled on the date, we had no speakers and no attendees. We knew that choosing the right group of speakers will attract attendees, so our next task was to invite such speakers. We decided to gather a mix of speakers from the academia and industry in order for attendees to be exposed not only to the newest advances undertaken at major research institutions, but also to understand how industrial leaders are utilizing big data and machine learning to provide services and products that have become essential parts of our daily lives.

Sooner than expected, the summer school date has arrived. Over the course of three days, the event featured lecturers from Carnegie Mellon University, the University of Illinois at Urbana-Champaign, Johns Hopkins University, Purdue University, the University of Maryland, the University of Toronto and the University of Pittsburgh. Lecturers from ANSYS, Rockwell Automation, Google and IBM were also in attendance and discussed the topic from an industry perspective. For the event program and all details about talks, please visit the official website for the event [1]. These talks covered a variety of different topics ranging from crowdsourced recording at scales, the role of big data and machine learning for real-world imaging and printing problems all the way to talks describing information forensics and deep learning.

As per attendees, over 100 registered attendees enjoyed the talks. While most of attendees were from the United States and Canada, we also had attendees from outside the North America. Furthermore, about 80% of attendees were IEEE student members, while the other 20% of attendees were senior researchers and engineers from universities and companies.

CONCLUSIONS

This was the first IEEE Signal Processing Society’s summer school in the United States, and the second one in North America. Based on the feedback received from attendees, we can definitely claim that the event was a success and the attendees were very satisfied with speakers and discussions that took place during breaks between talks. As for us, we gained a valuable experience in organizing an IEEE event, and we sincerely hope to have a chance to host another meeting at the University of Pittsburgh.

ACKNOWLEDGMENTS

We would like to express our endless gratitude for all speakers. They have made the event very valuable for all attendees.
We sincerely thank Jenna Berardino and Monica Bell from the University of Pittsburgh for organizing all logistical details around this event. Without their essential help and organizational skills, the event would not be as successful.
Last, but not least, we thank all volunteers and members of the organizing committee for their time and support.

REFERENCES:
[1] http://www.engineering.pitt.edu/IEEESPS/

Dr. Vikas Sindhwani from Google describing real-time learning and inferences on mobile systems

Dr. Andrew Moore from Carnegie Mellon University describing the role of big data and machine learning in search engines

Dr. Min Wu from the University of Maryland data analytics methods for information forensics

A group photo after the last talk at the summer school