Multilevel Training of Binary Morphological Operators

The design of binary morphological operators that are translation-invariant and locally defined by a finite neighborhood window corresponds to the problem of designing Boolean functions. As in any supervised classification problem, morphological operators designed from a training sample also suffer from overfitting. Large neighborhood tends to lead to performance degradation of the designed operator. This work proposes a multilevel design approach to deal with the issue of designing large neighborhood-based operators. The main idea is inspired by stacked generalization (a multilevel classifier design approach) and consists of, at each training level, combining the outcomes of the previous level operators. The final operator is a multilevel operator that ultimately depends on a larger neighborhood than of the individual operators that have been combined. Experimental results show that two-level operators obtained by combining operators designed on subwindows of a large window consistently outperform the single-level operators designed on the full window. They also show that iterating two-level operators is an effective multilevel approach to obtain better results.

Knowledge of firefighters with special paramedic training of the emergency management of avulsed teeth

Immediate replantation into the socket is the ideal procedure in cases of accidental avulsion of permanent teeth. In Brazil, firefighters with special paramedic training are in charge of providing first-aid care to victims of road accidents and might have to deal with tooth avulsions. This study assessed the knowledge of firefighters regarding the emergency management of avulsed teeth. Information was collected from a questionnaire submitted to 110 volunteer firefighters in seven cities in the São Paulo State (Brazil). The results revealed that 70.9% of the respondents did not know what tooth avulsion was; 53.6% did not know what tooth replantation was or defined it incorrectly; 60% would not act properly in tooth avulsion cases; 20.9% did not consider replantation of the avulsed tooth into the socket as a treatment option; the ideal time interval for tooth replantation was unknown to 40% of the interviewees; 90% of the participants answered that they would not be able to perform tooth replantation. Among those who considered themselves unable to perform tooth replantation, 47.3% chose saline as the best storage medium for an avulsed tooth, 21.8% chose milk, 3.6% chose the patient's mouth and 20% reported not knowing where to store the tooth; 81.8% of the firefighters reported not to have ever received any specific directions on tooth replantation and 100% of them considered this knowledge a requirement for first-aid care to accident victims. In conclusion, the knowledge of the surveyed firefighters regarding emergency management after tooth avulsion was unsatisfactory in several aspects that are important for the success of replantation procedures. Firefighters with special paramedic training should be educated on how to proceed in cases of dentoalveolar traumas and tooth avulsions in order to improve treatment prognosis and increase the survival rate of replanted teeth.

La formación de recursos humanos en el desarrollo económico y social de América Latina = The training of human resources in the economic and social development of Latin America

Incluye Bibliografía

A path- and label-cost propagation approach to speedup the training of the optimum-path forest classifier

In general, pattern recognition techniques require a high computational burden for learning the discriminating functions that are responsible to separate samples from distinct classes. As such, there are several studies that make effort to employ machine learning algorithms in the context of big data classification problems. The research on this area ranges from Graphics Processing Units-based implementations to mathematical optimizations, being the main drawback of the former approaches to be dependent on the graphic video card. Here, we propose an architecture-independent optimization approach for the optimum-path forest (OPF) classifier, that is designed using a theoretical formulation that relates the minimum spanning tree with the minimum spanning forest generated by the OPF over the training dataset. The experiments have shown that the approach proposed can be faster than the traditional one in five public datasets, being also as accurate as the original OPF. (C) 2014 Elsevier B. V. All rights reserved.