By Amrita Prasad (ZHAW)
It’s already been a month since we met as the Swiss Data Science community at our 3rd Swiss Conference on Data Science (SDS|2016), pushed again by ZHAW’s Datalab group and presented by SAP Switzerland.
Several additional organisations sponsored and supported the conference to give it a successful execution – the organising committee thanks IT Logix & Microsoft, PwC, Google, Zühlke, SGAICO, Hasler Stiftung and the Swiss Alliance for Data-Intensive Services for their support in bringing together a successful event!
This being the 3rd in the SDS conference series, we saw a tremendous rise in the number of participants as well as interest for presentation opportunities. The program opened with a welcome note by Hans Peter Gränicher, member of the board of the newly formed Swiss Association for Data-Intensive Services (Data+Service), which was already co-organizing this event. He spoke a few minutes about the vision and mission of the association. To find more information on the Data+Service association, please feel free to visit its official webpage.
The first keynote talk was presented by Prof. David Hand, Professor Emeritus of Mathematics at the Imperial College London. David Hand gave an insightful talk about the pitfalls of selection bias generated because of missing data and other reasons. He spoke of several examples where selection bias gave rise to wrong and/or dangerous assumptions in the data collected from an incomplete sample set. To view this and most other talks and slides, please visit the conference website.
The conference then presented an array of interesting keynote and technology oriented talks. Academics as well as industry experts presented their findings side by side to one another and discussed about the challenges and benefits these findings and technologies bring about.
Talking about current technologies and trends, there was a particularly interesting talk by Prof. Sven Crone (video to be released), from the Lancaster Centre for Forecasting, who talked about the hype and business reality in data science: While buzzwords come and go on Gartner’s hype cycle, he showed that today’s applications in forecasting are still deeply rooted in techniques more than 50 years old. In his hilarious way, he encouraged the audience to take the risk of additionally adopting also more modern algorithms. One key impression from his talk was that despite tremendous research and advances in current forecasting algorithms like the neural networks, the most frequently used algorithms were still age old like exponential smoothing and averages (algorithms from 1950s to 1960s).
Previously, Robert Rudolph from the board of Swissmem and Industrie2025 had addressed us as the genuinely data-savvy community with the plea to help the mechanical engineering industry make it through the digitization. He presented use cases showing the vast opportunities for analytics in the manufacturing business, basically inviting everybody to join.
The conference ended with a final keynote talk by Prof. Jürgen Schmidhuber from the Swiss AI lab, IDSIA. His talk focused on the history and evolution of AI as enabled by recurrent neural networks (RNNs), and the protagonists behind the field. He spoke about deep supervised / unsupervised / reinforcement learning, and discussed the latest state of the art results in numerous applications. The talk was positively received as an energetic end to the day long conference, specially highlighting Prof. Schmidhuber’s witty oration and keeping the audience gripped for more facts in AI.
The program was divided into 3 simultaneously ongoing tracks, and about 230 people joined in the conference. The job-wall was again very well received by companies putting up their job openings as well as CVs of people looking for a position in data science. There were booths from sponsors, with one particular highlight having a VR version of “Asteroids” to play live. We received very positive feedback after the conference, with some grumbling over food. We promise to improve!
A particular thing we liked about this conference was the generally high quality of the talks. The program offered something for each individual background, and provided very well received food for thought. A very warm “thank you” goes to all the session chairs and speakers who managed to keep the allotted time so well that the entire day felt smooth and peaceful, without being hectic, but filled with enthusiasm and energy.
We are looking forward to SDS|2017, that will be organized by the Swiss Alliance for Data-Intensive Services as an umbrella for everyone to get engaged. This is an open and serious call to get involved: We are looking for committed people and organizations on all levels, with all amounts of resources in time or money, be it in the organizing or program committee, as sponsors or speakers. Please help to spread the word, and please get in touch if you have questions or an idea how to get engaged. We look forward to hear from you and see you again next year!
Following are some more glimpses of the conference – enjoy!
Events such as these are a great way to stay up to date with the latest news and trends in the field of Data Science. Thank you for making this review it will be very helpful for people who did not get the chance to make it to this event.