BIG DATA International perspectives on the components of the data science skill set Marcel Worring Associate Director Amsterdam Datascience Informatics Institute & Amsterdam Business School University of Amsterdam DATA SCIENCE Amsterdam Data Science Goal of ADS It all started in November 2013 Amsterdam Data Science accelerates data science research by connecting, sharing and showcasing world-class technology, expertise and talent from the Amsterdam region on a regional, national and international level. 1
And the ecosystem develops Topic wise this is how it started 2
The AAA-program The true ecosystem Where we are Five knowledge institutes as official partner and counting AAA funding program Seven faculties Fifteen postdoc/phd positions and counting Over 2M in additional NWO funding and counting Nine non-academic and counting Outreach team and creating several labs 3
Our view on data science Decision Theory Visual Understand and Decide Business Research Distributed Processing Security Privacy Data Reasoning Knowledge representation Large Scale Databases Store and Process Governance Analyze and Model Multimedia Software Eng. System / Network Eng. Information Machine Learning Modeling and simulation Evaluation Criteria in Data Science Evaluation criteria Decision Theory Visual Understand and Decide Business Security Privacy Distributed Processing Large Scale Databases Speed Store and Process Efficiency SoftwareScalability Eng. System / Network Eng. Conformance Data Governance Information Reasoning Knowledge representation Multimedia Analyze Precision and Model Recall Model fit Modeling and simulation Machine Learning Evaluation Criteria in Data Science Decision Theory Visual Insight Understand and Decide Business Distributed Processing Large Scale Databases Speed Store and Process Efficiency Conformance Data Reasoning Knowledge representation Multimedia Analyze Precision and Model Recall The purpose of computing is insight, not numbers. Richard Hamming 1962 SoftwareScalability Eng. System / Network Eng. Information Model fit Modeling and simulation Machine Learning 4
Insight What is insight? Complex Insight is complex, involving all or large amounts of the given data in a synergistic way, not simply individual data values. Deep Insight builds up over time, accumulating and building on itself to create depth often generating further questions and, hence, further insight. North CG&A, 2006 Insight Relevant Insight is deeply embedded in the data domain, connecting the data to existing domain knowledge and giving it relevant meaning going beyond dry data analysis, to relevant domain impact. Unexpected Insight is often unpredictable, serendipitous, and creative. Qualitative Insight is not exact, can be uncertain and subjective, and can have multiple levels of resolution. [North 2008] Evaluation criteria Data Science Ecosystem Specialization is much easier than having wisdom Fudamental research Use-inspired research Applied Research Fundamental Research Grants TOP VIDI VENI LABS + STW NWO.. Companies Organizations Education Bachelor Education Education Master Track Data Science (IS) Artificial Intelligence Track Big Data Engineering (CS) Business SNE Econometrics: Track Big Data Post-Graduate MBA Big Data and Business Business Academy 5
Artificial Intelligence and Data Science BIG DATA ENGINEERING Advanced courses Part of regular Computer Science Program Machine Learning 2 Computer Vision 2 Natural Language Processing 2 Information 2 Deep Learning Track core courses: Large-Scale Data Engineering Web Data Processing Systems Web Services and Cloud-based Systems Information Visualization Data Mining Techniques Data Mining Techniques Information Visualization Probabilistic Robotics MBA INFORMATION STUDIES Data Science Education Ecosystem Fundamental Use-inspired Applied AI Computer Science Data Science Domain Specific Programs 6
Education Bachelor Education Master Track Data Science (IS) Artificial Intelligence Track Big Data Engineering (CS) Business SNE Econometrics: Track Big Data Post-Graduate MBA Big Data and Business Business Academy Amsterdam School of Data Science National Embedding International Embedding Insight center SOME CENTRES Why so many new centers? Data Science Center (FUTURE) IMPACT COLLABORATION FUNDING FROM THE UNIVERSITIES, GOVERNMENT, COMPANIES ECOSYSTEMS 7
EU Funding Priorities in funding Measurable The European Dilemma High impact for specific domains Man on the moon projects Data Science Funding Insight is hard to measure Domain specific solutions don t touch the core of data science We haven t been able to define our man on the moon goal EU collaboration Information Data Science Education Ecosystem Every country has everything Visual Security Decision Theory Business Privacy Reasoning Knowledge representation Multimedia Software Large Scale Databases Eng. Modeling and simulation Governance System / Network Eng. Machine Learning Distributed Processing Specialized Fundamental research AI Computer Science Use-inspired research Data Science Specialized Applied Research Domain Specific Room should be made for The various data science principles and ecosystems and developing the science in data science Fundamental Use-inspired Applied Data science specialization Summary Conclusion 8
Amsterdam Data Science 9