SBC
BRACIS 2018 - IBM Research | São Paulo, Brazil
Tutorials

Introduction to Deep Learning –Theory and Practice

  • Dr. Aparecido Nilceu Marana
  • Associate Professor at São Paulo State University (UNESP)
  • Gustavo Botelho de Souza
  • PhD candidate in Comp.Science Federal University of São Carlos (UFSCar)
  • Dr. João Paulo Papa
  • Associate Professor at São Paulo State University (UNESP)

Language: portuguese

Abstract: Recently, deep learning architectures emerged as good alternatives for solving complex problems and have reached state-of-the-art results in many tasks such as Computer Vision, Natural Language Processing, etc., due to their power of abstraction and robustness, working with abstract and high-level features, self-learned from the training data. The main goal of this tutorial is to introduce the attendees to the deep neural networks, i.e., their architectures, training algorithms and activation functions,especially the Convolutional Neural Networks (CNNs), as well as to present some deep learning tools and implement examples using the Caffeframework, motivating themto further researches

Maximizing your reward at BRACIS: an introduction to reinforcement learning

  • Anderson Rocha Tavares
  • UFMG
  • Bruno Castro da Silva
  • UFRGS
  • Luiz Chaimowicz
  • UFMG
  • Renato L. F. Cunha
  • UFMG and IBM Research

Language: portuguese

Abstract: Inspired by the way humans and other animals learn basic behaviors in nature, Rein-forcement Learning (RL) is concerned with learning by interacting with the environment.Basically, an agent learns the actions it should perform at a given state (it learns apolicy)by exploring the state space and receiving rewards for its actions. This intuitive approachhas gained much attention in recent years, with applications ranging from robotics to digi-tal games. Thus, the objective of this tutorial is to give an overview of this important area,presenting the basic concepts, the main algorithms and discussing some applications.

Slides:

Find the presentation slides here.