A frequency-domain approach to improve ANNs generalization quality via proper initialization.

The ability to train a network without memorizing the input/output data, thereby allowing a good predictive performance when applied to unseen data, is paramount in ANN applications. In this paper, we propose a frequency-domain approach to evaluate the network initialization in terms of quality of training, i.e., generalization capabilities. As an alternative to the conventional time-domain methods, the proposed approach eliminates the approximate nature of network validation using an excess of unseen data. The benefits of the proposed approach are demonstrated using two numerical examples, where two trained networks performed similarly on the training and the validation data sets, yet they revealed a significant difference in prediction accuracy when tested using a different data set. This observation is of utmost importance in modeling applications requiring a high degree of accuracy. The efficiency of the proposed approach is further demonstrated on a real-world problem, where unlike other initialization methods, a more conclusive assessment of generalization is achieved. On the practical front, subtle methodological and implementational facets are addressed to ensure reproducibility and pinpoint the limitations of the proposed approach.

The capacity of glutathione reductase in cell protection from the toxic effect of heated oils.

This empirical study tries to focus on the evidence that the wrong use of oil in food cooking leads to health problems. High temperatures associated with the repeated use of the same oil lead to the breakdown of some fatty acids, forming numerous toxic polymer compounds and peroxides. The obtained data have showed that the ratios of polymer compounds reached 11.3% in oil heated continuously for 10 h at 220 degrees C, and 37.8% in frying oil (FO). Moreover, the polar compound ratios reached 25.6% and 47.6% in continuously heated oil (CHO) and FO, respectively. However, the peroxide concentrations were 157.1 and 133.6 mM/kg in CHOs and FOs, respectively. The observed results have allowed the study of the role of the glutathione redox system in the detoxification and elimination of different toxic peroxides resulting from heated oils. On a diet of 10% of CHO and FO, a significant increase in glutathione peroxidase (GPx) and glutathione reductase (GR) activities appears. This combined relationship between the decreased glutathione content and the increased GPx and GR activities in rats fed on CHO and FO confirms the participation of the glutathione redox system in the detoxifying reactions of continuously accumulated peroxides.

Urinary markers of workers chronically exposed to mercury vapor.

The effects of chronic mercury intoxication on urinary markers in workers from northeast Algeria were investigated. Workmen were chosen from highly and moderately mercury-exposed factories, while controls were selected from a nonexposed site. The number of proteinuria cases were higher in the highly exposed subjects, although the nature (glomerular or tubular) of proteinuria remains unclear. However, it appears difficult to assess the degree of renal disturbance among the different exposure levels, such as the amount of excreted proteins, which have not clearly reflected the kidney lesion severity. The results also reveal that urine acidity increased progressively with increased levels of exposure, while a remarkable inverse relationship between urinary pH and urinary Hg in the highly exposed workers has been recorded. Furthermore, the significant differences in blood and urinary mercury concentrations of the three sites reflect the dose-response relationships.

The toxicological effects of the herbicide 2,4-DCPA on progesterone levels and mortality in Wistar female rats.

The effect of the ester butylglucol of 2,4-Dichlorophenoxyacetic acid (herbicide) on the rate of fertilization and on progesterone levels during gestation in rats, has been studied. The experiments consisted of four groups; one receiving the drug in the drinking water at a concentration of 125 mg/kg/day (the agriculture dose), the second and the third group received per os doses of 50 and 100 mg/kg/day of the herbicide, while the fourth group served as controls. The obtained results show that all treatments lead to a decrease in both indexes of females' fertility and in that of newborn mortality. On the other hand, the index of fertilization has been highly affected by the drug application, with a clear delay in fertilization in animals treated with 100 mg/kg/day and 125 mg/kg/day. Progesterone levels have also been suppressed in individuals receiving the herbicide at 50 and 125 mg/kg/day.