From January 23 to 27, Fundação Getulio Vargas’ School of Applied Mathematics (FGV EMAp) is holding a Summer School on Data Science. This in-person event will bring together students, professors and researchers to discuss technical aspects of advances in the theory and application of data science.

The event’s program includes four minicourses and 12 lectures. The activities will take place daily, from 9 am to 6:30 pm, at the FGV Cultural Center (Praia de Botafogo, 186, Rio de Janeiro).

Minicourses:

Introduction to Streaming Algorithms

• Professor: Nivan Ferreira (Pernambuco Federal University)
• Date and time: January 23 to 25, 9 am to 11 am

Streaming algorithms generally employ probabilistic methods to compute summaries for a stream of data. The aim of this course is to present the main ideas behind algorithms designed to compute statistical summaries for data streams, as well as uses of these algorithms to analyze large data sets.

AI for Earth Observation

• Professor: Laura Rosa (Wageningen University)
• Date and time: January 23 to 25, 11:30 am to 1:30 pm

The free availability of a growing body of Earth observation (EO) imagery has created an urgent need to develop effective and efficient algorithms capable of processing large amounts of data and extracting crucial information for various applications, from environmental monitoring to precision agriculture. In this context, deep learning algorithms have proven to be effective at learning relevant features directly from the data and they are used in many EO image analysis tasks, including land use and land cover classification, object detection, change detection and domain adaptation. This minicourse aims to introduce students to the evolution of computer vision applied to EO data, focusing on innovative deep learning-based models and their application to real-world problems.

From Algebraic Topology to Data Analysis

• Professor: Raphaël Tinarrage (FGV)
• Date and time: January 25 to 27, 9 am to 11 am, and January 27, 3 pm to 5 pm

This course will introduce topological data analysis (TDA) and, in particular, persistent homology. In TDA, we seek to discover and understand the topology (that is, the shape) of datasets. Instead of applying rigid models to data, we preserve its inherent complexity and explore it through topological invariants. By illuminating data analysis from a new angle, TDA opens the door to new insights and discoveries.

The Art of Modeling via Gaussian Processes

• Professor: César Mattos (Ceará Federal University)
• Date and time: January 25 to 27, 11:30 am to 1:30 pm, and January 27, 5:30 pm to 7:30 pm

Gaussian processes (GPs) are non-parametric Bayesian models that allow the quantification of uncertainty in the context of probabilistic machine learning. This minicourse introduces fundamental principles and some more modern developments in modeling via GPs. The purpose of the meetings is to inspire participants to incorporate GP models into their applications and to encourage study and research in this area.

For more information about the minicourses and to see the complete program, visit the Summer School website.