Single-preference consensus measures based on models of ecological evenness

We investigate the relationship between consensus measures used in different settings depending on how voters or experts express their preferences. We propose some new models for single-preference voting, which we derive from the evenness concept in ecology, and show that some of these can be placed within the framework of existing consensus measures using the discrete distance. Finally, we suggest some generalizations of the single-preference consensus measures allowing the incorporation of more general notions of distance.

Neuron’s spikes noise level classification using hidden markov models

Considering that the uncertainty noise produced the decline in the quality of collected neural signal, this paper proposes a signal quality assessment method for neural signal. The method makes an automated measure to detect the noise levels in neural signal. Hidden Markov Models were used to build a classification model that classifies the neural spikes based on the noise level associated with the signal. This neural quality assessment measure will help doctors and researchers to focus on the patterns in the signal that have high signal to noise ratio and carry more information.

Regularizing topic discovery in emrs with side information by using hierarchical bayesian models

We propose a novel hierarchical Bayesian framework, word-distance-dependent Chinese restaurant franchise (wd-dCRF) for topic discovery from a document corpus regularized by side information in the form of word-to-word relations, with an application on Electronic Medical Records (EMRs). Typically, a EMRs dataset consists of several patients (documents) and each patient contains many diagnosis codes (words). We exploit the side information available in the form of a semantic tree structure among the diagnosis codes for semantically-coherent disease topic discovery. We introduce novel functions to compute word-to-word distances when side information is available in the form of tree structures. We derive an efficient inference method for the wddCRF using MCMC technique. We evaluate on a real world medical dataset consisting of about 1000 patients with PolyVascular disease. Compared with the popular topic analysis tool, hierarchical Dirichlet process (HDP), our model discovers topics which are superior in terms of both qualitative and quantitative measures.

Exploiting side information in Bayesian nonparametric models and their applications

 My research is to exploit side information into advanced Bayesian nonparametric models. We have developed some novel models for data clustering and medical data analysis and also have made our methods scalable for large-scale data. I have published my research in several journal and conference papers.

Small-variance asymptotics for bayesian nonparametric models with constraints

The users often have additional knowledge when Bayesian nonparametric models (BNP) are employed, e.g. for clustering there may be prior knowledge that some of the data instances should be in the same cluster (must-link constraint) or in different clusters (cannot-link constraint), and similarly for topic modeling some words should be grouped together or separately because of an underlying semantic. This can be achieved by imposing appropriate sampling probabilities based on such constraints. However, the traditional inference technique of BNP models via Gibbs sampling is time consuming and is not scalable for large data. Variational approximations are faster but many times they do not offer good solutions. Addressing this we present a small-variance asymptotic analysis of the MAP estimates of BNP models with constraints. We derive the objective function for Dirichlet process mixture model with constraints and devise a simple and efficient K-means type algorithm. We further extend the small-variance analysis to hierarchical BNP models with constraints and devise a similar simple objective function. Experiments on synthetic and real data sets demonstrate the efficiency and effectiveness of our algorithms.

Are CGE Models still useful in economic policy making?

This paper develops background considerations to help better framing the results of a CGE exercise. Three main criticisms are usually addressed to CGE efforts. First, they are too aggregate, their conclusions failing to shed light on relevant sectors or issues. Second, they imply huge data requirements. Timeliness is frequently jeopardised by out-dated sources, benchmarks referring to realities gone by. Finally, results are meaningless, as they answer wrong or ill-posed questions. Modelling demands end up by creating a rather artificial context, where the original questions lose content. In spite of a positive outlook on the first two, crucial questions lie in the third point. After elaborating such questions, and trying to answer some, the text argues that CGE models can come closer to reality. If their use is still scarce to give way to a fruitful symbiosis between negotiations and simulation results, they remain the only available technique providing a global, inter-related way of capturing economy-wide effects of several different policies. International organisations can play a major role supporting and encouraging improvements. They are also uniquely positioned to enhance information and data sharing, as well as putting people from various origins together, to share their experiences. A serious and complex homework is however required, to correct, at least, the most dangerous present shortcomings of the technique.

Weather-based prediction of anthracnose severity using artificial neural network models

Data were collected and analysed from seven field sites in Australia, Brazil and Colombia on weather conditions and the severity of anthracnose disease of the tropical pasture legume Stylosanthes scabra caused by Colletotrichum gloeosporioides. Disease severity and weather data were analysed using artificial neural network (ANN) models developed using data from some or all field sites in Australia and/or South America to predict severity at other sites. Three series of models were developed using different weather summaries. of these, ANN models with weather for the day of disease assessment and the previous 24 h period had the highest prediction success, and models trained on data from all sites within one continent correctly predicted disease severity in the other continent on more than 75% of days; the overall prediction error was 21.9% for the Australian and 22.1% for the South American model. of the six cross-continent ANN models trained on pooled data for five sites from two continents to predict severity for the remaining sixth site, the model developed without data from Planaltina in Brazil was the most accurate, with >85% prediction success, and the model without Carimagua in Colombia was the least accurate, with only 54% success. In common with multiple regression models, moisture-related variables such as rain, leaf surface wetness and variables that influence moisture availability such as radiation and wind on the day of disease severity assessment or the day before assessment were the most important weather variables in all ANN models. A set of weights from the ANN models was used to calculate the overall risk of anthracnose for the various sites. Sites with high and low anthracnose risk are present in both continents, and weather conditions at centres of diversity in Brazil and Colombia do not appear to be more conducive than conditions in Australia to serious anthracnose development.

Evaluation and identification of lightning models by artificial neural networks

This paper describes a novel approach for mapping lightning models using artificial neural networks. The networks acts as identifier of structural features of the lightning models so that output parameters can be estimated and generalized from an input parameter set. Simulation examples are presented to validate the proposed approach. More specifically, the neural networks are used to compute electrical field intensity and critical disruptive voltage taking into account several atmospheric and structural factors, such as pressure, temperature, humidity, distance between phases, height of bus bars, and wave forms. A comparative analysis with other approaches is also provided to illustrate this new methodology.

Information and mechanical models of intelligence: What can we learn from cognitive science?

The impact of new advanced technology on issues that concern meaningful information and its relation to studies of intelligence constitutes the main topic of the present paper. The advantages, disadvantages and implications of the synthetic methodology developed by cognitive scientists, according to which mechanical models of the mind, such as computer simulations or self-organizing robots, may provide good explanatory tools to investigate cognition, are discussed. A difficulty with this methodology is pointed out, namely the use of meaningless information to explain intelligent behavior that incorporates meaningful information. In this context, it is inquired what are the contributions of cognitive science to contemporary studies of intelligent behavior and how technology may play a role in the analysis of the relationships established by organisms in their natural and social environments. © John Benjamins Publishing Company.

Information-entropic measure of energy-degenerate kinks in two-field models

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Artificial neural networks applied to bandsawing process control

In the search for productivity increase, industry has invested on the development of intelligent, flexible and self-adjusting method, capable of controlling processes through the assistance of autonomous systems, independently whether they are hardware or software. Notwithstanding, simulating conventional computational techniques is rather challenging, regarding the complexity and non-linearity of the production systems. Compared to traditional models, the approach with Artificial Neural Networks (ANN) performs well as noise suppression and treatment of non-linear data. Therefore, the challenges in the wood industry justify the use of ANN as a tool for process improvement and, consequently, add value to the final product. Furthermore, Artificial Intelligence techniques such as Neuro-Fuzzy Networks (NFNs) have proven effective, since NFNs combine the ability to learn from previous examples and generalize the acquired information from the ANNs with the capacity of Fuzzy Logic to transform linguistic variables in rules.

Design-based random permutation models with auxiliary information

We extend the random permutation model to obtain the best linear unbiased estimator of a finite population mean accounting for auxiliary variables under simple random sampling without replacement (SRS) or stratified SRS. The proposed method provides a systematic design-based justification for well-known results involving common estimators derived under minimal assumptions that do not require specification of a functional relationship between the response and the auxiliary variables.