Mario Cantú-Sifuentes

Laboratory Development and Field Validation of Phormia regina (Diptera: Calliphoridae)

ABSTRACT Immature blow flies (Diptera: Calliphoridae) collected from decomposing human remains are often used to determine the minimum postmortem interval (PMImin). Phormia regina (Meigen) is a common blow fly of cosmopolitan distribution that is often associated in such cases. P. regina development at two different cyclic temperatures was examined in this study. A field validation study was conducted to determine the accuracy of applying these data to determine the PMImin. Minimal total development time was 32.52 d at cyclic 14.0 ±2.0°C and 16.60 d at cyclic 20.5 ±3.1°C. The minimal larval development was significantly different (P < 0.05) across temperatures. Larval development needed 15.5 d at 14.0°C and 7.5 d at 20.5°C. For the validation study, instar, mean, and maximum of length and weight data of the larvae collected in the field were analyzed with data generated from the 20.5°C treatment, as it more closely reflected the field conditions experienced. Accuracy in estimating PMImin, was highly variable depending on the unit of measurement used and instar of P. regina collected from the field. Using the oldest instar to estimate a PMImin resulted in ranges that always encompassed the true time of colonization. Accuracy in hours when using measurements units as mean length or weight, and maximal length or weight, varied among the larval instars. In the first instar the greatest overestimation was made with maximal weight while the greatest underestimation was made with mean weight. The most accurate estimate produced with first instars was based on maximal length. In the second instar, there was no overestimation and the greatest underestimation was made with mean weight and the most accurate estimate produced was with maximal length. In the third instar, the greatest overestimation was made with maximal length, and the greatest underestimation was made with mean weight. The estimated time of colonization based on maximal weight was most accurate for third instars.

Fuzzy reliability analysis with only censored data

Abstract It is expected that samples in reliability analysis contain both censored and complete failure data; thus, the maximum likelihood method is used to estimate the parameters of the related distribution. Nonetheless, samples may contain only censored data; therefore introducing a high degree of uncertainty which does result in non-viability for either the likelihood method or for statistical inference. This paper proposes the use of fuzzy probability theory to account for the uncertainty and the prior knowledge of the process in the parameters׳ estimation, for censored data. The proposed method was applied to risk based inspection. Results demonstrate that our method represents a reliable option for using the expert knowledge about the component and the physics of the failure mode. Additionally, an inspection time was estimated based on target risk; the results confirm that the methodology could be used to develop maintenance plans.

Optimization by Canonical Analysis in a Radial Basis Function

Abstract Generally, statistical methods and mathematical models are useful for process optimization. Nonetheless, other methods might be used for modeling and optimizing the manufacturing process. Among these, we can mention the neural networks and the Radial Basis Function technique. Hence, a suitable alternative is complementing statistical methods and neural networks as a Hybrid Learning Process. This work applies the Radial Basis Function Canonical Analysis in order to achieve the welding process optimization. One of the most important results is that the Radial Basis Function neural networks along with the Canonical Analysis are really useful methods. These methods are applied for predicting the optimal point, which establishes a reliable method for the process modeling and optimizing. The Canonical Analysis can determine stationary and saddle points, as it was in this case of study, which Canonical Analysis with RBF represented it adequately and can plot a surface and contour lines. Since in this case of study there is a surface that contains a ridge saddle system, also often called minimax. Then the results show that the Canonical Analysis can explore the region with oblique stationary and rising ridge systems. In this way, the RBF neural network with Canonical Analysis could be an alternative method for analyzing data, whenever the Hybrid Learning Process is adequate or satisfies the test assumption and fulfills the evaluation criteria. In this case of study, validation is represented by the Hybrid Learning Process (Radial Basis Function with Canonical Analysis) presenting an excellent effectiveness. As a conclusion we can say that the resulting Radial Basis Function has improved the model accuracy after using the Canonical Analysis.

Statistical inference in a redesigned Radial Basis Function neural network

A Hybrid Learning Process method was fitted into a RBF. The resulting redesigned RBF intends to show how to test if the statistical assumptions are fulfilled and to apply statistical inference to the redesigned RBFNN bearing in mind that it allows to determine the relationship between a response (to a process) and one or more independent variables, testing how much each factor contributes to the total variation of the response is also feasible. The results show that statistical methods such as inference, Residual Analysis, and statistical metrics are all good alternatives and excellent methods for validation of the effectiveness of the Neural Network models. The foremost conclusion is that the resulting redesigned Radial Basis Function improved the accuracy of the model after using a Hybrid Learning Process; moreover, the new model also validates the statistical assumptions for using statistical inference and statistical analysis, satisfying the assumptions required for ANOVA to determine the statistical significance and the relationship between variables.

Multivariate statistical inference in a radial basis function neural network

Abstract Regression models and analysis of variance are widely used methods for analyzing manufacturing processes in the industry. Its objective is to analyze the effect of process input variables on a final quality characteristic. However, some processes have complex data that cannot be described by linear regression or have several product quality characteristics to be controlled, having several correlated responses makes it difficult to analyze and optimize the process, thus, it is common to ignore the correlation between responses and make several independent models for each response, causing problems for decision-making based on these models. This paper shows the application of multivariate statistical analysis in a Radial Basis Function neural network, considering the statistical significance between independent and dependent variables and the correlation and verifying if the assumptions for this analysis are fulfilled. The results evidence that the multivariate statistical analysis in the Radial Basis Function is a good method to analyze the process based on correlated variables: it satisfies the assumptions required for this analysis, it determines the process variation, and it also determines the most important variable that had influence in the process used to evaluate the permanent mold casting process.