- - Fundamentals and Applications of Artificial and Computational Intelligence
Description:
Artificial and Computational Intelligence involves the study of mathematical models and computational tools inspired by natural and biological models for intelligent data manipulation, involving acquisition, representation, manipulation and classification of knowledge, through the ability to deduce or infer new knowledge about existing knowledge, such as new relationships involving facts and concepts, or emerging relationships from new events that may occur.
That is, from existing knowledge about data and its explicit and implicit relationships, formalized through logic and mathematics, models and methods are developed for the representation, manipulation, classification, reasoning, inference, analysis and decision-making about new examples of data, generally involving the solution of complex problems, often of a qualitative nature.
The theoretical bases of this line of research are based on data science, machine learning, deep learning, extreme learning, evolutionary networks, agent theory and multi-agent systems, fuzzy logic and its generalization, fusion, aggregation, overlap, grouping, and implication functions, fuzzy inference systems, interval mathematics, among others.
This line of research is particularly involved with models and tools aimed at explaining and treating uncertainty. To this end, it is based on the development of theories, models, techniques, methods and intelligent systems that offer solutions tolerant to subjectivity, imprecision, uncertainty, incomplete, partial or conflicting information, partial truth and partial possibility, which appear in the modeling of complex systems and/or imperfect information, so common in the current context of Science and Technology. One of the objectives of this line of research is to seek approximations or combinations of different theories, models, techniques and methods that are capable of adequately reflecting imperfection, achieving tractability, robustness, reliability and solutions at a reasonable cost.
Considering distributed artificial intelligence, this line of research considers the development of multi-agent systems as a viable alternative to the modeling and implementation of complex systems. The research carried out involves cognitive agent models, based on decision theories, and hybrids; models of organization of agent systems - organizations, institutions, societies of agents; models of coordination and regulation of interactions - exchange values, game and decision theory; models of societies of agents - social, cultural, economic, and political aspects; methodologies for specifying agent-based systems. The main applications studied include multi-robot systems, distributed manufacturing systems, models and methodologies for social and environmental simulation.
An important application for the computational models developed in this line is in the development of solutions for Bioinformatics. One of the major challenges lies precisely in the analysis of large volumes of complex data, from which reliable models of biological processes can be generated, such as those involved in the production of drugs. Theories, models, methods and techniques in molecular docking and virtual screening, data science and machine learning applied to biological data, molecular dynamics simulations, protein folding, genetic regulatory networks, and simulation of biological systems are proposed.
This line of research works in a multi and interdisciplinary approach, also considering applications in Interaction and Communication Technologies in Education, investigating educational and assistive technologies with the use of Artificial and Computational Intelligence. The following themes are considered: virtual and augmented reality in education; robotics in inclusive education; simulators and educational games.
- Faculty:
Adriano Velasque Werhli
Alessandro de Lima Bicho
Bruno Lopes Dalmazo
Eder Mateus Nunes Gonçalves
Eduardo Nunes Borges
Graçaliz Pereira Dimuro
Hélida Salles Santos
Karina dos Santos Machado
Leonardo Ramos Emmendorfer
Rafael Alceste Berri
Regina Barwaldt
Rodrigo Andrade de Bem
Tiago da Cruz Asmus
- Research Groups:
Computational Biology (Combi-L)
Flexible Computing (CompFlex)
Information Management (Ginfo)
Informatics in Education (InfoEduc)
Multi-Agent Systems (GPSMA)
- - Robotics, Intelligent Automation and Computing Systems
Description:
The main objective of this line of research is to study, develop and apply conceptual aspects of computing and computational tools to solve complex problems in automation and robotics, especially those associated with applications focused on the coastal and oceanic ecosystem. The term “intelligent”, used in the definition of the line, arises from the need to deal with the complexity in modeling solutions to be applied in oceanic and coastal scenarios, as well as the uncertainties, incompleteness, scalability and time and space restrictions typical of these systems. In addition, the line addresses topics in Computing Systems, with a main focus on Digital Systems, Embedded Systems, Operating Systems, Distributed Systems, Computer Networks and Security. Such systems are considered strategic within the Exact Sciences and Engineering since they are present in almost all industrialized products and large-scale online applications.
The aim is to study and develop new algorithms and computational techniques for the analysis, modeling and control of sensor and actuator networks, whether they are part of Automation Systems or grouped into Robotic Systems. More precisely, considering the application scenarios, the aim is to study the computational aspects involved in engineering projects that aim to build artifacts with distributed perception, decision-making and actuation capabilities, in real time, and autonomous, in real, virtual or mixed environments.
The aforementioned aspects mean that research in this area produces significant added value in the national and global economic scenario. Furthermore, mastering device design and manufacturing technologies and implementing large-scale systems are key elements for the development of industry and research in all areas of knowledge.
More specifically, we intend to research models, methods and techniques in: Computer Systems and Technologies for Perception - acquisition and processing of sensory information, models for representation and description of environments, recognition and interpretation of sensory information (mapping, localization, tracking); Computer Systems and Technologies for Decision Making - control, navigation, planning and scheduling systems; Architectures for Sensor and Actuator Networks - architectures for (multi) robots, automation systems (buses, RFIDs, SCADAs), manufacturing systems, ubiquitous systems; Mixed Environments - virtual and mixed reality, teleoperation, simulation and synthesis of environments, dual worlds (telepresence and teleoperation), internet of things, and reconstruction; Design of micro and nano systems; Development of tools to aid in the design of integrated circuits; Design of digital systems; Design of high energy efficiency systems; Design of fault-tolerant systems; Assessment of the reliability of nanotechnologies; Cybersecurity; Infrastructure and Internet interconnection; Parallel and distributed programming; Distributed systems; Replication and fault tolerance in distributed systems; and Development and verification of concurrent systems.
- Faculty:
Alessandro de Lima Bicho
Eder Mateus Nunes Gonçalves
Eduardo Nunes Borges
Emanuel da Silva Diaz Estrada
Marcelo Rita Pias
Nelson Lopes Duarte Filho
Paulo Francisco Butzen
Paulo Lilles Jorge Drews Junior
Pedro de Botelho Marcos
Rafael Alceste Berri
Rafael Budim Schvittz
Rodrigo Andrade de Bem
Rodrigo da Silva Guerra
Silvia Silva da Costa Botelho
Vagner Santos da Rosa
Vinicius Garcia Pinto
Vinicius Menezes de Oliveira
- Research Groups: