Two seminars on data science
- Dr. Carlos Soares
- donderdag 31 maart 2016
Niels Bohrweg 1
2333 CA Leiden
Visiting scholar dr. Carlos Soares from the University of Porto will be giving two seminars on data science, which are open to all interested parties. No registration is required.
Recent advances in metalearning: from metafeatures to streams (13:00-13:50)
Metalearning consists of applying machine learning approaches to obtain models that relate the characteristics of problems with the performance of methods. Despite the potential, the metalearning field is still at an early stage of its development, with many challenges, including problem characterization, metadata collection and application in new areas, such as recommendation systems, streaming data and sensor networks. In this talk we will discuss some of the latest developments regarding these challenges.
Prescriptive analytics: 2 case studies in retail and health industries (14:00-14:50)
Prescriptive analytics combines predictive models with other decision support techniques to provide systems that address broader problems. While descriptive and predictive analytics can help you characterize the sales of a product in a store or predict its future values, predictive analytics can be used to define distribution of the shelf space that will optimize the overall sales of the store. In this talk we discuss two applications: shelf space allocation for a large, multi-store retailer and surgery scheduling in a large hospital.