Neurofuzzy design and model construction of nonlinear dynamical processes from data

A common problem in many data based modelling algorithms such as associative memory networks is the problem of the curse of dimensionality. In this paper, a new two-stage neurofuzzy system design and construction algorithm (NeuDeC) for nonlinear dynamical processes is introduced to effectively tackle this problem. A new simple preprocessing method is initially derived and applied to reduce the rule base, followed by a fine model detection process based on the reduced rule set by using forward orthogonal least squares model structure detection. In both stages, new A-optimality experimental design-based criteria we used. In the preprocessing stage, a lower bound of the A-optimality design criterion is derived and applied as a subset selection metric, but in the later stage, the A-optimality design criterion is incorporated into a new composite cost function that minimises model prediction error as well as penalises the model parameter variance. The utilisation of NeuDeC leads to unbiased model parameters with low parameter variance and the additional benefit of a parsimonious model structure. Numerical examples are included to demonstrate the effectiveness of this new modelling approach for high dimensional inputs.

Comparison of TAMSAT and CPC rainfall estimates with raingauges, for southern Africa

Two different TAMSAT (Tropical Applications of Meteorological Satellites) methods of rainfall estimation were developed for northern and southern Africa, based on Meteosat images. These two methods were used to make rainfall estimates for the southern rainy season from October 1995 to April 1996. Estimates produced by both TAMSAT methods and estimates produced by the CPC (Climate Prediction Center) method were then compared with kriged data from over 800 raingauges in southern Africa. This shows that operational TAMSAT estimates are better over plateau regions, with 59% of estimates within one standard error (s.e.) of the kriged rainfall. Over mountainous regions the CPC approach is generally better, although all methods underestimate and give only 40% of estimates within 1 s.e. The two TAMSAT methods show little difference across a whole season, but when looked at in detail the northern method gives unsatisfactory calibrations. The CPC method does have significant overall improvements by building in real-time raingauge data, but only where sufficient raingauges are available.

Expected value, reward outcome, and temporal difference error representations in a probabilistic decision task

In probabilistic decision tasks, an expected value (EV) of a choice is calculated, and after the choice has been made, this can be updated based on a temporal difference (TD) prediction error between the EV and the reward magnitude (RM) obtained. The EV is measured as the probability of obtaining a reward x RM. To understand the contribution of different brain areas to these decision-making processes, functional magnetic resonance imaging activations related to EV versus RM (or outcome) were measured in a probabilistic decision task. Activations in the medial orbitofrontal cortex were correlated with both RM and with EV and confirmed in a conjunction analysis to extend toward the pregenual cingulate cortex. From these representations, TD reward prediction errors could be produced. Activations in areas that receive from the orbitofrontal cortex including the ventral striatum, midbrain, and inferior frontal gyrus were correlated with the TD error. Activations in the anterior insula were correlated negatively with EV, occurring when low reward outcomes were expected, and also with the uncertainty of the reward, implicating this region in basic and crucial decision-making parameters, low expected outcomes, and uncertainty.

Prediction of metabolisable energy concentrations of fresh cut grass using digestibility data measured with non-pregnant non-lactating cows

Pasture-based ruminant production systems are common in certain areas of the world, but energy evaluation in grazing cattle is performed with equations developed, in their majority, with sheep or cattle fed total mixed rations. The aim of the current study was to develop predictions of metabolisable energy (ME) concentrations in fresh-cut grass offered to non-pregnant non-lactating cows at maintenance energy level, which may be more suitable for grazing cattle. Data were collected from three digestibility trials performed over consecutive grazing seasons. In order to cover a range of commercial conditions and data availability in pasture-based systems, thirty-eight equations for the prediction of energy concentrations and ratios were developed. An internal validation was performed for all equations and also for existing predictions of grass ME. Prediction error for ME using nutrient digestibility was lowest when gross energy (GE) or organic matter digestibilities were used as sole predictors, while the addition of grass nutrient contents reduced the difference between predicted and actual values, and explained more variation. Addition of N, GE and diethyl ether extract (EE) contents improved accuracy when digestible organic matter in DM was the primary predictor. When digestible energy was the primary explanatory variable, prediction error was relatively low, but addition of water-soluble carbohydrates, EE and acid-detergent fibre contents of grass decreased prediction error. Equations developed in the current study showed lower prediction errors when compared with those of existing equations, and may thus allow for an improved prediction of ME in practice, which is critical for the sustainability of pasture-based systems.

Analysis Scheme in the Ensemble Kalman Filter

This paper discusses an important issue related to the implementation and interpretation of the analysis scheme in the ensemble Kalman filter . I t i s shown that the obser vations must be treated as random variables at the analysis steps. That is, one should add random perturbations with the correct statistics to the obser vations and generate an ensemble of obser vations that then is used in updating the ensemble of model states. T raditionally , this has not been done in previous applications of the ensemble Kalman filter and, as will be shown, this has resulted in an updated ensemble with a variance that is too low . This simple modification of the analysis scheme results in a completely consistent approach if the covariance of the ensemble of model states is interpreted as the prediction error covariance, and there are no further requirements on the ensemble Kalman filter method, except for the use of an ensemble of sufficient size. Thus, there is a unique correspondence between the error statistics from the ensemble Kalman filter and the standard Kalman filter approach

Cherry picking with synthetic controls

The synthetic control (SC) method has been recently proposed as an alternative method to estimate treatment e ects in comparative case studies. Abadie et al. [2010] and Abadie et al. [2015] argue that one of the advantages of the SC method is that it imposes a data-driven process to select the comparison units, providing more transparency and less discretionary power to the researcher. However, an important limitation of the SC method is that it does not provide clear guidance on the choice of predictor variables used to estimate the SC weights. We show that such lack of speci c guidances provides signi cant opportunities for the researcher to search for speci cations with statistically signi cant results, undermining one of the main advantages of the method. Considering six alternative speci cations commonly used in SC applications, we calculate in Monte Carlo simulations the probability of nding a statistically signi cant result at 5% in at least one speci cation. We nd that this probability can be as high as 13% (23% for a 10% signi cance test) when there are 12 pre-intervention periods and decay slowly with the number of pre-intervention periods. With 230 pre-intervention periods, this probability is still around 10% (18% for a 10% signi cance test). We show that the speci cation that uses the average pre-treatment outcome values to estimate the weights performed particularly bad in our simulations. However, the speci cation-searching problem remains relevant even when we do not consider this speci cation. We also show that this speci cation-searching problem is relevant in simulations with real datasets looking at placebo interventions in the Current Population Survey (CPS). In order to mitigate this problem, we propose a criterion to select among SC di erent speci cations based on the prediction error of each speci cations in placebo estimations

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.

Damage detection in nonlinear structures using discrete-time volterra series

Structural damage identification is basically a nonlinear phenomenon; however, nonlinear procedures are not used currently in practical applications due to the complexity and difficulty for implementation of such techniques. Therefore, the development of techniques that consider the nonlinear behavior of structures for damage detection is a research of major importance since nonlinear dynamical effects can be erroneously treated as damage in the structure by classical metrics. This paper proposes the discrete-time Volterra series for modeling the nonlinear convolution between the input and output signals in a benchmark nonlinear system. The prediction error of the model in an unknown structural condition is compared with the values of the reference structure in healthy condition for evaluating the method of damage detection. Since the Volterra series separate the response of the system in linear and nonlinear contributions, these indexes are used to show the importance of considering the nonlinear behavior of the structure. The paper concludes pointing out the main advantages and drawbacks of this damage detection methodology. © (2013) Trans Tech Publications.

Proposta de construção de um amortecedor de vibração ajustável, TVA, utilizando fluido magnetoreológico

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Projeto otimizado de redes neurais artificiais para predição da rugosidade em processos de usinagem com a utilização da metodologia de projeto de experimentos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Avaliação cefalométrica comparativa da previsibilidade manual e computadorizada dos resultados do tratamento ortodôntico-cirúrgico bimaxilar em pacientes face longa: estudo retrospectivo

Pós-graduação em Odontologia - FOA

Estudo cefalométrico comparativo dos traçados predictivos e resultados ortodôntico-cirúrgicos em pacientes com face longa

The purpose of this research was to compare, by cephalometric analysis (McNamara and Legan & Burstone) the predictive tracings (by methods manual, and by softwares Dentofacial Planner Plus and Dolphin Image) with the post surgical results. Were selected the pre and post surgical lateral telerradiograph (six months after orthognatic surgery) of the 25 long face patients treated with combined orthognatic surgery. Were made the prediction tracings for each method and comparing cephalometrically with the post surgical results. This protocol was repeated once more for the error method evaluation, and the statistical was made by variance analysis and Tuckey overtest. The results show more frequency of the cephalometric values' aproximation of the post surgical results when the manual method (50% of the similarity with the post surgical result), followed of the DFPlus (31,2%) and Dolphin (18,8%) softwares. The experimental condition permits to conclude that the manual method had more precision, although the previsibility of the digital methods was reasonable satisfactory.

Comparação entre modelos de análise genômica utilizando dados simulados e dados reais em ovinos

Pós-graduação em Genética e Melhoramento Animal - FCAV

Accuracy, precision and robustness of different methods to obtain samples from silages in fermentation studies

The objective of this study was to evaluate accuracy, precision and robustness of two methods to obtain silage samples, in comparison with extraction of liquor by manual screw-press. Wet brewery residue alone or combined with soybean hulls and citrus pulp were ensiled in laboratory silos. Liquor was extracted by a manual screw-press and a 2-mL aliquot was fixed with 0.4 mL formic acid. Two 10-g silage samples from each silo were diluted in 20 mL deionized water or 17% formic acid solution (alternative methods). Aliquots obtained by the three methods were used to determine the silage contents of fermentation end-products. The accuracy of the alternative methods was evaluated by comparing mean bias of estimates obtained by manual screw-press and by alternative methods, whereas precision was assessed by the root mean square prediction error and the residual error. Robustness was determined by studying the interaction between bias and chemical components, pH, in vitro dry matter digestibility (IVDMD) and buffer capacity. The 17% formic acid method was more accurate for estimating acetic, butyric and lactic acids, although it resulted in low overestimates of propionic acid and underestimates of ethanol. The deionized water method overestimated acetic and propionic acids and slightly underestimated ethanol. The 17% formic acid method was more precise than deionized water for estimating all organic acids and ethanol. The robustness of each method with respect to variation in the silage chemical composition, IVDMD and pH is dependent on the fermentation end-product at evaluation. The robustness of the alternative methods seems to be critical at the determination of lactic acid and ethanol contents.

Evaluating Prediction Rules for t-Year Survivors With Censored Regression Models

Suppose that we are interested in establishing simple, but reliable rules for predicting future t-year survivors via censored regression models. In this article, we present inference procedures for evaluating such binary classification rules based on various prediction precision measures quantified by the overall misclassification rate, sensitivity and specificity, and positive and negative predictive values. Specifically, under various working models we derive consistent estimators for the above measures via substitution and cross validation estimation procedures. Furthermore, we provide large sample approximations to the distributions of these nonsmooth estimators without assuming that the working model is correctly specified. Confidence intervals, for example, for the difference of the precision measures between two competing rules can then be constructed. All the proposals are illustrated with two real examples and their finite sample properties are evaluated via a simulation study.