Theory meets applications in healthcare, public administration and business
Recent computational and data accessibility developments have made it possible to transform raw data into useful predictions and detailed descriptions of the world. For those who deal with real-world data, the problems of large amounts of heterogeneous, noisy, and incomplete datasets come hand-in-hand with the benefits of ubiquitous, up-to-date, and rich information. I will present some of the work I have done with my group with such data, interfacing with domain experts from healthcare, public administration, and businesses. These efforts exposed scientific challenges to the data analysis tools available today. Such challenges fuelled my research, and I will illustrate how principled theory matched real-world applications. Tools from statistical data analysis, to numerical and optimization techniques, are enablers for the research developments that can meet those challenges.
21/04/2021 13:30
Knowledge-Based Systems
Claudia Soares is an Assistant Professor at NOVA School of Science and Technology, and a researcher at NOVA LINCS, Portugal. She uses real-world data problems to identify the shortcomings of current machine learning, data science, and Big Data methods. She applies optimization, statistics, and probability theory to address those gaps, developing robust, interpretable, and fair learning methods that can be trusted in real life. Her application areas are in healthcare, transportation, environmental and urban sciences, and space. Prof. Soares holds a degree in modern languages and literature, and a B.Sc., M.Sc., and Ph.D. in Electrical and Computer Engineering.
