This post is by Modulos’ head of data science, Dr. Anna Weigel.
At the end of January the Modulos data science team attended the 2020 Applied Machine Learning Days (AMLD) at EPFL. AMLD is a unique event as it is neither a classic trade show nor a pure science conference: it showcases the most recent break throughs in research, as well as the state of the art developments in industry.
AMLD has been growing steadily over the past few years and this year had 2’500 participants. In addition to the plenary talks, 29 separate tracks covered topics from AI and aviation, cities, climate, pharma, and nutrition, to imaging, education, governance, and space.
Many of the talks touched upon issues related to data collection and data management in the context of distributed business processes. Another prominent topic was the often high failure rate of data science projects and what differentiates them from classical IT projects.
Conversations with other data scientists also showed the need to be able to estimate in advance how well a machine learning solution is likely to perform.
My personal highlights include the talks by Prof. Michal Kosinski and Edward Snowden on privacy, as well as Prof. David Autor’s keynote on the impact of AI on the labor market. I also greatly enjoyed our team dinner which involved a variety of crêpes in Lausanne’s lovely old town and am looking forward to returning for AMLD 2021.