Connectionist simulation of quantification skills

The study of numerical abilities, and how they are acquired, is being used to explore the continuity between ontogenesis and environmental learning. One technique that proves useful in this exploration is the artificial simulation of numerical abilities with neural networks, using different learning paradigms to explore development. A neural network simulation of subitization, sometimes referred to as visual enumeration, and of counting, a recurrent operation, has been developed using the so-called multi-net architecture. Our numerical ability simulations use two or more neural networks combining supervised and unsupervised learning techniques to model subitization and counting. Subitization has been simulated using networks employing unsupervised self-organizing learning, the results of which agree with infant subitization experiments and are comparable with supervised neural network simulations of subitization reported in the literature. Counting has been simulated using a multi-net system of supervised static and recurrent backpropagation networks that learn their individual tasks within an unsupervised, competitive framework. The developmental profile of the counting simulation shows similarities to that of children learning to count and demonstrates how neural networks can learn how to be combined together in a process modelling development.     

Testing Forecast Accuracy of Foreign Exchange Rates: Predictions from Feed Forward and Various Recurrent Neural Network Architectures

In this research, we work with data of futures contracts on foreign exchange rates for British pound (BP), Canadian dollar (CD), and Japanese yen (JY) that are traded at the Chicago Mercantile Exchange (CME) against US dollars. We model relationships between exchange rates in these currencies using linear models, feed forward artificial neural networks (ANN), and three versions of recurrent neural networks (RNN1, RNN2 and RNN3) for predicting exchange rates in these currencies against the US dollar. Our results on forecast evaluations based on AGS test the tests of forecast equivalence between any two competing models among the entire models employed for each of the series show that ANN and the three versions of RNN models offer superior forecasts for predicting BP, CD and JY exchange rates although the forecast evaluations based on MGN test are in sharp contrast. On the other hand forecast based on SIGN test shows that ANN and all the versions of RNN models offer superior forecasts for BP and CD in exception of JY exchange rates. The results for forecast evaluation for all the models for each of the series based on summary measures of forecast evaluations show that RNN3 model appears to offer the most accurate predictions of BP and RNN1 for JP exchange rates. However, none of the RNN models appear to be statistically superior to the benchmark (i.e., linear model) for predicting CD exchange rates.      

Numerical solution of second‐order,two‐point boundary value problems using continuous genetic algorithms

Second‐order, two‐point boundary‐value problems are encountered in many engineering applications including the study of beam deflections, heat flow, and various dynamic systems. Two classical numerical techniques are widely used in the engineering community for the solution of such problems; the shooting method and finite difference method. These methods are suited for linear problems. However, when solving the non‐linear problems, these methods require some major modifications that include the use of some root‐finding technique. Furthermore, they require the use of other basic numerical techniques in order to obtain the solution. In this paper, the author introduces a novel method based on continuous genetic algorithms for numerically approximating a solution to this problem. The new method has the following characteristics; first, it does not require any modification while switching from the linear to the non‐linear case; as a result, it is of versatile nature. Second, this approach does not resort to more advanced mathematical tools and is thus easily accepted in the engineering application field. Third, the proposed methodology has an implicit parallel nature which points to its implementation on parallel machines. However, being a variant of the finite difference scheme with truncation error of the order O(h²), the method provides solutions with moderate accuracy. Numerical examples presented in the paper illustrate the applicability and generality of the proposed method. Copyright © 2004 John Wiley & Sons, Ltd.     

Storage and permeation of hydrogen molecule,atom and ions (H+ and H−) through silicon carbide nanotube; a DFT approach

The barriers for the encapsulation and decapsulation of hydrogen ions (cationic hydrogen and hydride), atom, and molecule through silicon carbide nanotube are thoroughly studied. DFT method is selected to measure the kinetic barriers for the passage of hydrogen atom, ions and molecule through nanotube via scanning potential energy surface. The kinetic barriers for the passage (encapsulation and decapsulation) of hydrogen are very important to understand the mechanism of hydrogen storage and release. The barriers for the permeation of H, H⁺ and H^? across SiC nanosheet are lower compared to hydrogen molecule (H₂). The exohedral and endohedral adsorption of hydrogen ions (cation and anion), atom and exohedral hydrogen molecule on silicon carbide are exothermic in nature. Whereas the encapsulation of hydrogen molecule in silicon carbide is endothermic. Electronic properties are analyzed through measurement of energy gap between highest occupied and lowest unoccupied molecular orbitals gap (G_H-L) and the density of state (DOS) spectra. The G_H-L analysis reveals that endohedral complexes have more pronounced effect on electronic properties compared to exohedral complexes. The SiC nanotube has highly favorable properties for storage and release of hydrogen ions, and atom.     

The multiaxial weld root fatigue of butt welded joints subjected to uniaxial loading

In this study, the fatigue strength of inclined butt welds subjected to a proportional multiaxial stress state generated by uniaxial loading is studied in nominal and local stress concepts. The local methodologies studied included principal stress hypothesis, von Mises stress hypothesis and modified Wöhler curve method. Nominal methodologies included modified Gough–Pollard interaction equation, the design equation in Eurocode3 and the interaction equation in DNV standard. Results are evaluated along with data published in relevant literature. It is observed that both local and nominal stress assessment methods are able to estimate multiaxial fatigue strength. No obvious difference in fatigue strength is observed in the nominal stress concept, but the notch stress concept is able to capture a decrease in fatigue strength in shear‐dominated joints. It is concluded that modified Wöhler curve method is a suitable tool for the evaluation of fatigue strength in joints dominated by both normal and shear stresses.     

Differences between t‐norms in fuzzy control

The application of conjunctive aggregation functions in fuzzy control systems with n inputs is discussed, and the effect of the choice of a continuous t‐norm in the inference phase for Takagi–Sugeno–Kang (TSK) systems is computed. A continuous t‐norm modeling AND connective in antecedent part of fuzzy rules can be reduced just to strict or nilpotent t‐norm. The isomorphism of strict (nilpotent) t‐norms enables simpler fitting of TSK fuzzy system parameters and reduces the computational complexity. Similar principle can also be used in the case of some noncommutative conjunctive aggregation functions modeling AND connective. The effect of the choice of a continuous t‐norm is then evaluated on well‐known case studies in fuzzy control, the Sinc function, and the urban traffic noise control system.     

Bacteriological Evaluation of Gingival Crevicular Fluid in Teeth Restored Using Fixed Dental Prostheses: An In Vivo Study

The present in vivo study determined the microbiological counts of the gingival crevicular fluid (GCF) among patients with fixed dental prostheses fabricated using three different techniques. A total of 129 subjects were divided into three study groups: first, cobalt-chrome-based, metal-ceramic prostheses fabricated by the conventional method (MC, n = 35); the second group consisted of cobalt-chrome-based, metal-ceramic prostheses fabricated by the computer-aided design and computer-aided manufacturing (CAD/CAM) technique (CC-MC, n = 35); the third group comprised zirconia-based ceramic prostheses fabricated using the CAD/CAM technique (CC-Zr, n = 35). The control consisted of 24 patients using prostheses fabricated with either MC, CC-MC, or CC-Zr. The GCF was obtained from the subjects before treatment, and 6 and 12 months after the prosthetic treatment. Bacteriological and bacterioscopic analysis of the GCF was performed to analyze the patients’ GCF. The data were analyzed using SPSS V20 (IBM Company, Chicago, IL, USA). The number of microorganisms of the gingival crevicular fluid in all groups at 12 months of prosthetic treatment reduced dramatically compared with the data obtained before prosthetic treatment. Inflammatory processes in the periodontium occurred slowly in the case of zirconium oxide-based ceramic constructions due to their biocompatibility with the mucous membranes and tissues of the oral cavity as well as a reduced risk of dental biofilm formation. This should be considered by dentists and prosthodontists when choosing restoration materials for subjects with periodontal pathology.     

T‐norms in subtractive clustering and backpropagation

The tuning of a fuzzy model is discussed in the context of choices made between different t‐norms. The effects of the choice is illustrated by looking at two fuzzy models initially generated, respectively, by grid partition and a novel variant of subtractive clustering. The new variant of subtractive clustering introduced in the paper is based on the standard method of subtractive clustering, where in this new method, the measure of similarity and thus also cluster shapes depend on a choice of t‐norm T. © 2010 Wiley Periodicals, Inc.     

Fatigue assessment of high‐frequency mechanical impact (HFMI)‐treated welded joints subjected to high mean stresses and spectrum loading

The fatigue strength of welded joints can be improved with various post‐weld treatment methods. High‐frequency mechanical impact treatment is a residual stress modification technique that creates compressive residual stresses at the weld toe. However, these beneficial residual stresses may relax under certain loading conditions. In this paper, previously published fatigue data for butt and fillet welded joints subjected to high stress ratios and variable amplitude cyclic stresses were evaluated in relation to the current International Institute of Welding (IIW) recommendations on fatigue strength improvement and a proposed IIW design guideline for high‐frequency mechanical impact‐treated welded joints. The evaluation showed that the current IIW recommendations resulted in both non‐conservative and overly conservative fatigue strength estimations depending on the applied stress level, whereas the proposed fatigue assessment guideline fitted the current data well.