Claudio Fuentes

Limitations of the Log-Logistic Model for the Analysis of Sigmoidal Microbial Inactivation Data for High-Pressure Processing (HPP)

This study identified limitations of the log-logistic model to evaluate microbial inactivation kinetics by high-pressure processing (HPP) including the need to assign a numerical value to “approximate” the undefined expression log10t = 0 and the misinterpretation of its parameters due to a derivation flaw. Peer-reviewed HPP microbial inactivation data were adjusted to a sigmoidal equation (SIG), the original “vitalistic” log-logistic models (VIT-1, VIT-6), and two functions that did not follow the original derivation procedure (LOG-1, LOG-6). Their goodness of fit was determined utilizing the coefficient of determination (R2) and Akaike information criteria (AIC). The shape of the survival curve greatly influenced the performance of log-logistic models. VIT and LOG models performed equally when the kinetic curve showed a sigmoidal shape, and the numerical values of their parameter estimates were identical regardless of the log10 (t = 0) approximation. Conversely, most concave curves yielded inaccurate parameter estimates for all models. LOG-1 and VIT-1 performed best when log10t = 0 was −1 or −2, whereas LOG-6 and VIT-6 yielded best results for values of −3 to −9. SIG ranked last for most datasets but occasionally performed best (Akaike weight factor wAICi = 0.40–1.00) when microbial survival counts showed clear sigmoidal shapes. VIT models consistently displayed R2 ≥ 0.98, and their parameters can be interpreted within a “biological” context using the corrected derivation shown for LOG models. However, concave curves are more frequently observed for HPP microbial inactivation, and fitting the experimental data to log-logistic models deems unnecessary.

Offline training for improving online performance of a genetic algorithm based optimization model for hourly multi-reservoir operation

A novel framework, which incorporates implicit stochastic optimization (Monte Carlo method), cluster analysis (machine learning algorithm), and Karhunen-Loeve expansion (dimension reduction technique) is proposed. The framework aims to train a Genetic Algorithm-based optimization model with synthetic and/or historical data) in an offline environment in order to develop a transformed model for the online optimization (i.e., real-time optimization). The primary output from the offline training is a stochastic representation of the decision variables that are constituted by a series of orthogonal functions with undetermined random coefficients. This representation preserves covariance structure of the simulated decisions from the offline training as gains some knowledge regarding the search space. Due to this gained knowledge, better candidate solutions can be generated and hence, the optimal solutions can be obtained faster. The feasibility of the approach is demonstrated with a case study for optimizing hourly operation of a ten-reservoir system during a two-week period. A model training framework incorporate Monte Carlo method, machine learning algorithm and dimension reduction technique.Trained model significantly improve the online performance of optimization.Generic representation allow a broad application in environmental and water resources system.

Marginal likelihood estimation of negative binomial parameters with applications to RNA-seq data

RNA-Seq data characteristically exhibits large variances, which need to be appropriately accounted for in any proposed model. We first explore the effects of this variability on the maximum likelihood estimator (MLE) of the dispersion parameter of the negative binomial distribution, and propose instead to use an estimator obtained via maximization of the marginal likelihood in a conjugate Bayesian framework. We show, via simulation studies, that the marginal MLE can better control this variation and produce a more stable and reliable estimator. We then formulate a conjugate Bayesian hierarchical model, and use this new estimator to propose a Bayesian hypothesis test to detect differentially expressed genes in RNA-Seq data. We use numerical studies to show that our much simpler approach is competitive with other negative binomial based procedures, and we use a real data set to illustrate the implementation and flexibility of the procedure.

The Role of Functional Categories in L2 Spanish: Persistence of L1 CP Values in IL

Abstract English irrealis modality is usually captured by syntactic structures that often exclude overt complementizers that and for and include indicative or infinitival morphology. The Spanish equivalents to English structures of irrealis modality, however, typically require the overt subordinate marker que (‘that’) followed by a [+finite] subordinate verb. These syntactic differences between English and Spanish produce asymmetrical surface structures caused by functional categories (FCs). The primary goal of this study is to examine the role of FCs, and more specifically, complementizers (CPs), in the syntactic development of L1 English speakers with L2 Spanish. The results of our data suggest that L1 CP properties (i.e. CP = Ø) persisted in learners’ interlanguage (IL) systems, following the work of Schwartz & Sprouse (1996). As a result, learners at all levels of instruction intuitively perceived English-like [-overt] CP structures as more grammatical than their [+overt] CP counterparts, even at advanced levels of acquisition.

Biomarker validation with an imperfect reference: Issues and bounds

Motivated by the goal of evaluating a biomarker for acute kidney injury, we consider the problem of assessing operating characteristics for a new biomarker when a true gold standard for disease status is unavailable. In this case, the biomarker is typically compared to another imperfect reference test, and this comparison is used to estimate the performance of the new biomarker. However, errors made by the reference test can bias assessment of the new test. Analysis methods like latent class analysis have been proposed to address this issue, generally employing some strong and unverifiable assumptions regarding the relationship between the new biomarker and the reference test. We investigate the conditional independence assumption that is present in many such approaches and show that for a given set of observed data, conditional independence is only possible for a restricted range of disease prevalence values. We explore the information content of the comparison between the new biomarker and the reference test, and give bounds for the true sensitivity and specificity of the new test when operating characteristics for the reference test are known. We demonstrate that in some cases these bounds may be tight enough to provide useful information, but in other cases these bounds may be quite wide.

Error-rate estimation in discriminant analysis of non-linear longitudinal data: A comparison of resampling methods.

Consider longitudinal observations across different subjects such that the underlying distribution is determined by a non-linear mixed-effects model. In this context, we look at the misclassification error rate for allocating future subjects using cross-validation, bootstrap algorithms (parametric bootstrap, leave-one-out, .632 and . 632 + ), and bootstrap cross-validation (which combines the first two approaches), and conduct a numerical study to compare the performance of the different methods. The simulation and comparisons in this study are motivated by real observations from a pregnancy study in which one of the main objectives is to predict normal versus abnormal pregnancy outcomes based on information gathered at early stages. Since in this type of studies it is not uncommon to have insufficient data to simultaneously solve the classification problem and estimate the misclassification error rate, we put special attention to situations when only a small sample size is available. We discuss how the misclassification error rate estimates may be affected by the sample size in terms of variability and bias, and examine conditions under which the misclassification error rate estimates perform reasonably well.

Predicting pregnancy outcomes using longitudinal information: a penalized splines mixed-effects model approach

We propose a semiparametric nonlinear mixed-effects model (SNMM) using penalized splines to classify longitudinal data and improve the prediction of a binary outcome. The work is motivated by a study in which different hormone levels were measured during the early stages of pregnancy, and the challenge is using this information to predict normal versus abnormal pregnancy outcomes. The aim of this paper is to compare models and estimation strategies on the basis of alternative formulations of SNMMs depending on the characteristics of the data set under consideration. For our motivating example, we address the classification problem using a particular case of the SNMM in which the parameter space has a finite dimensional component (fixed effects and variance components) and an infinite dimensional component (unknown function) that need to be estimated. The nonparametric component of the model is estimated using penalized splines. For the parametric component, we compare the advantages of using random effects versus direct modeling of the correlation structure of the errors. Numerical studies show that our approach improves over other existing methods for the analysis of this type of data. Furthermore, the results obtained using our method support the idea that explicit modeling of the serial correlation of the error term improves the prediction accuracy with respect to a model with random effects, but independent errors. Copyright

Analysis of Vibrio vulnificus Infection Risk When Consuming Depurated Raw Oysters.

A beta Poisson dose-response model for Vibrio vulnificus food poisoning cases leading to septicemia was used to evaluate the effect of depuration at 15 °C on the estimated health risk associated with raw oyster consumption. Statistical variability sources included V. vulnificus level at harvest, time and temperature during harvest and transportation to processing plants, decimal reductions (SV) observed during experimental circulation depuration treatments, refrigerated storage time before consumption, oyster size, and number of oysters per consumption event. Although reaching nondetectable V. vulnificus levels (<30 most probable number per gram) throughout the year and a 3.52 SV were estimated not possible at the 95% confidence level, depuration for 1, 2, 3, and 4 days would reduce the warm season (June through September) risk from 2,669 cases to 558, 93, 38, and 47 cases per 100 million consumption events, respectively. At the 95% confidence level, 47 and 16 h of depuration would reduce the warm and transition season (April through May and October through November) risk, respectively, to 100 cases per 100 million consumption events, which is assumed to be an acceptable risk; 1 case per 100 million events would be the risk when consuming untreated raw oysters in the cold season (December through March).