Ribeirão Preto, SP, Brazil
Invited Speakers
 |
» DeLiang Wang, PhD Talk Title: Cocktail party processing |
 |
» Frank Dignum, PhD
Talk Title: Organizing software agents |
 |
» José Carlos Príncipe, PhD |
 |
» Pedro Domingos, PhD Dept. of Computer Science & Engineering University of Washington USA http://www.cs.washington.edu/homes/pedrod/ Talk Title: Unifying Logical and Statistical AI |
 |
» Tom Mitchell, PhD Director, Center for Automated Learning and Discovery Fredkin Professor of Computer Science Carnegie Mellon University USA http://www.cs.cmu.edu/~tom/ Talk Title: Machine Learning and Analyzing Human Brain Activity Abstract: In recent years there has been a breakthrough in instruments for observing human brain activity, and even more recently machine learning methods have emerged as a valuable new approach to analyzing this data. This talk will present our recent research exploring human brain activity associated with the meanings of different words and pictures. For example, using machine learning methods we have trained classifiers capable of decoding whether a person is reading a word about tools or buildings, independent of whether the word is presented in English or Portuguese. We will present these results and examine the machine learning methods used. We will also present our recent research on Hidden Process Models: an approach to learning probabilistic time series models capable of modeling and tracking multiple cognitive processes that overlap in time and space within the brain. |
 |
» Peter A. Flach, PhD Talk Title: Reinventing Machine Learning with ROC Analysis Abstract: I Receiver Operating Characteristics (ROC) analysis originated from signal detection theory, as a model of how well a receiver is able to detect a signal in the presence of noise. Its key feature is the distinction between hit rate (or true positive rate) and false alarm rate (or false positive rate) as two separate performance measures. It has been introduced to machine learning relatively recently, in response to classification tasks with varying class distributions or misclassification costs. ROC analysis is set to cause a paradigm shift in machine learning. Separating performance on classes is almost always a good idea from an analytical perspective. In this talk I will mainly illustrate how ROC analysis can be used to understand fundamental algorithms such as the separate-and-conquer or sequential covering rule learning algorithm, by tracing its trajectory through a sequence of ROC spaces. |
Workshop Invited Speakers
TIL'2006
 |
» Gladis Maria de Barcellos Almeida
Talk Title:
The role of Computer Science in Linguistic research: expanding the horizon of language description Abstract: The advances of studies in the areas of Computational Linguistics and Linguistics of Corpus and the consequent improvement and development of Portuguese Natural Language Processing (NLP) tools start to interfere directly with the way of making linguistic research. At the same time where there are favorable and reliable conditions for linguistic description, the proper linguistic description, to have success, finishes demanding from researchers new linguists studies, since the possibility to deal with huge corpora allows to the description of recurrent linguistic phenomena that had been previously difficult or impossible to perceive, given that the procedures of description counted only on manual resources. In Brazil, where the advances of the Computational Linguistics and Corpus Linguistics delayed happen in relation to the countries of the hemisphere north, it is necessary to reconsider the relation between Computer Science and Linguistics. The Linguistics is not only benefited of the Computer Science, but it needs it. I intend, therefore, to present this new scenery for the linguistic research in Brazil, illustrating it with results from recent projects developed in the Studies and Research in Terminology Group (GETerm-UFSCar) and the Interinstitutional Center for Research and Development in Computational Linguistics (NILC-USP/São Carlos). I will focus on compilation of corpora, creation of text typologies and automatic terminology extraction for the generation of dictionaries. |
 |
» Isaac Bejar Talk Title: Educational assessment systems in the 21st century: The increasing importance of a multidisciplinary and global perspective Abstract: The need to improve educational systems to be more competitive to the demands of the knowledge economy is fueling an increase in the use of assessments worldwide. In this presentation I will discuss recent developments assessment design that facilitate identifying, and therefore designing, optimally the various "moving parts" of an assessment. Specifically, I will discuss the need to approach the design of assessments with a global and multidisciplinary perspective. That approach makes it possible, with some planning, to repurpose or reuse assessment designs in a different locality. The talent needed for designing such assessments seldom resides in one locality suggesting that assessment design will become increasingly multidisciplinary and global. The presentation will illustrate these trends with projects ranging from the licensing of architects to educational assessments for accountability. The use of automation for item generation and scoring will be emphasized as key methodological and technical enablers of cost effective and valid assessments. |
IBERAGENTS'2006
| » Antônio Carlos da Rocha Costa Universidade Católica de Pelotas, UCPEL Brazil Talk Title: An exchange value-based foundation for the ethics of artificial agents | |
 |
» Jaime Simão Sichman Universidade de São Paulo, USP Brazil Talk Title: Organization Oriented Programming: From Closed to Open MAS |
Partner Sponsored Talk
 |
» SAS Talk Title: Programa Acadêmico e a Solução SAS para Data Mining A Prática de Ensino do SAS dedica-se a criar uma cultura de liderança, legitimidade e responsabilidade na educação - ao fornecer aos professores, administradores, estudantes, as melhores soluções para alinhar os recursos com resultados baseados nas evidências.
|
|