03/05/2023: Fotos do aniversário do PIPGEs!
O nosso 8º dia temático foi dedicado à comemoração dos 10 anos do PIPGEs.
Nesta ocasião, recebemos 4 palestrantes brasileiros de grande projeção nacional e internacional, e 4 egressos do PIPGEs (de mestrado e de doutorado).
O evento ocorreu no dia 28/04, acontecendo no CINA-UFSCar na parte da manhã (9h-12h) e no Auditório Prof. Luiz Antonio Favaro do ICMC-USP na parte da tarde (14h-18h30).
Contou com a participação de aproximadamente 50 pessoas! Neste link estão algumas fotos (quem fez mais, por favor compartilhar).
Abaixo segue a programação que tivemos durante o dia!
Programação:
9h-10h: Mesa de abertura
Além de alguns membros do PIPGEs, teremos o prazer de contar nesta mesa com:
- representação da pró-reitoria de PPG da UFSCar,- representação da diretoria do ICMC-USP,
- representação da diretoria do CCET-UFSCar.
10h-10h50: Bias Correction in Clustered Underreported Data
Profa. Rosangela Loschi (UFMG)
11h-11h50: Robust beta regression through the logit transformation
Profa. Silvia Ferrari (USP)
Beta regression models are employed to model continuous response variables in the unit interval, like rates, percentages, or proportions. Their applications rise in several areas, such as medicine, environment research, finance, and natural sciences. The maximum likelihood estimation is widely used to make inferences for the parameters. Nonetheless, it is well-known that the maximum likelihood-based inference suffers from the lack of robustness in the presence of outliers. Such a case can bring severe bias and misleading conclusions. Recently, robust estimators for beta regression models were presented in the literature. However, these estimators require non-trivial restrictions in the parameter space, which limit their application. This paper develops new robust estimators that overcome this drawback. Their asymptotic and robustness properties are studied, and robust Wald-type tests are introduced. Simulation results evidence the merits of the new robust estimators. Inference and diagnostics using the new estimators are illustrated in an application to health insurance coverage data.
12h-14h: Almoço
14h-16h: Apresentações curtas de egressos do PIPGEs
Análise de textos por meio de processos estocásticos narepresentação word2vec (Gabriela Massoni, IME- USP)
Contributions to item response theory and cognitive diagnostic models (Prof. Marcelo A. da Silva, ESALQ)
Uncertainty in Recommender Systems (Victor A. Coscrato, UCC-Irlanda)
Existência e comparações entre espectros L^q de medidas estacionárias, tempos de espera e tempos de retorno (Prof. Vitor G. de Amorim, IFSP-Araraquara)
16h-16h30: Café
16h30-17h20: The unreasonable effectiveness of data science
Prof. Renato Assunção, UFMG
There are three factors responsible for the revolution brought about by artificial intelligence: (1) the constant increase in computational capacity; (2) the accumulation of large amounts of data generating insights and enabling the creation of data-driven products; (3) the development of statistical learning theory and its algorithms. The alignment of these planets allowed great success in difficult tasks such as the development of virtual assistants and chatbots, self-driving car, the automatic translation between languages, and the early detection of unspecified anomalies in vital signs. In this talk, I will present an overview of these developments from a historical point of view focusing on the contribution brought by Statistics. I will illustrate this presentation with examples from my own research on epidemiological surveillance using social media data, space-time demographic forecasting, and the Bayesian spatial partitioning of space-time maps.
17h30-18h20: Detecting Renewal States in texts of Brazilian and European Portuguese
Profa. Nancy L. Garcia, UNICAMP
In a previous work, Galves et al (2012) analyzed the differences between written texts of Brazilian and European Portuguese. Their method was based on Markov chains with variable length having a renewal state in order to perform bootstrap. Markov chains with variable length are useful parsimonious stochastic models able to model sequences of discrete symbols. The suffixes of the past that are relevant to predict the future symbol are called contexts.
Sometimes a single state is a context, and looking at the past and finding this specific state makes the further past irrelevant. States with such property are called renewal states and they can be used to split the chain into independent and identically distributed blocks. In order to evaluate the hypothesis that some particular state is a renewal state, we propose the use of the Intrinsic Bayes Factor.