Design of a fast convergent backpropagation algorithm based on optimal control theory

The main contribution of this paper is using optimal control theory for improving the convergence rate of backpropagation algorithm. In the proposed approach, the learning algorithm of backpropagation is modeled as a minimum time control problem in which the step-size of its learning factor is considered as the input of this model. In contrast to the traditional backpropagation, learning algorithms which select the step-size by trial and error, it is selected adaptively based on optimal control criterion. The effectiveness of the proposed algorithm is evaluated in two simulations: XOR and 3-bit parity. In both simulation examples, the proposed algorithm outperforms well in speed and the ability to escape from local minima.     

Effect of Manganese on Radiation Vulcanization of Natural Rubber

Manganese was added as a promoter to investigate physico-mechanical properties of radiation-vulcanized natural rubber latex (RVNRL) films. RVNRL films were prepared by the addition of Mn with the concentration range 0–30 ppm to natural rubber latex and irradiated with various radiation doses (0–20 kGy). Tensile strength, tear strength, and cross-linking density of the irradiated rubber films increased with increasing the concentration of Mn ions as well as radiation doses. In contrast, elongation at break, permanent set, and swelling ratio of the films were decreased under the same conditions. The concentration of Mn ions and radiation doses were optimized and found to be 20 ppm and 12 kGy, respectively. The maximum tensile and tear strengths of irradiated rubber films were observed as 29.12 MPa and 44.78 N/mm, respectively at the optimum conditions. The mechanical properties of the films increased markedly with the addition of Mn until they attained the highest values of 33.88 MPa and 54.77 N/mm, respectively. These enhancements, which reached approximately 20% at the most favorable conditions, can be explained by the effect of transition metals in view of Fajan’s rules regarding the covalent character of ionic bonds and suggest that the higher the difference in charges between cation and anion, the higher the ability to form distortion or polarization of ions.     

Comparative study of the alignment method on experimental and simulated chromatographic data

One of the major problems in the signal comparison of chromatographic data is the variability of response caused by instrumental drifts and others instabilities. Measures of quality control and evaluation of conformity are inherently sensitive to shift. It is essential to be able to compare test samples to reference samples in an evolutionary analytical environment by offsetting the inevitable drift. Therefore, prior to any multivariate analysis, the alignment of analytical signals is a compulsory preprocessing step. During recent years, many researchers have taken a greater interest in the study of the alignment. The present paper is an updated review on the alignment algorithms, methods, and improvements used in chromatography. The study is dedicated to one‐dimensional signals. Several of the exposed methods have common theoretical bases and can differ through their optimization methods. The main issue for the operator is to choose the appropriate method according to the type of signals to be processed.     

Electrophoretic deposition of PVA coated hydroxyapatite on 316L stainless steel

Polyvinyl alcohol (PVA) coated hydroxyapatite was deposited onto a 316L stainless steel substrate by electrophoretic deposition. Deposition was carried out in a methanol suspension at pH 5.5 using a graphite rod as an anode. Parameters such as PVA concentration, deposition voltage and time were optimized to achieve a homogeneous, crack-free adhesive coating. Techniques such as X-ray diffraction and Fourier transform infrared spectroscopy were used to study the phase composition of the coated materials and the stability of hydroxyapatite in the presence of PVA.     

Papain Loaded Poly(ε-Caprolactone) Nanoparticles<Emphasis Type=Italic>: In-silico and</Emphasis><Emphasis Type=Italic>In-Vitro</Emphasis> Studies

Papain is a protease enzyme with therapeutic properties that are very valuable for medical applications. Poly(ε-caprolactone) (PCL) is an ideal polymeric carrier for controlled drug delivery systems due to its low biodegradability and its high biocompatibility. In this study, the three-dimensional structure and action mechanism of papain were investigated by in vitro and in silico experiments using molecular dynamics (MD) and molecular docking methods to elucidate biological functions. The results showed that the size of papain-loaded PCL nanoparticles (NPs) and the polydispersity index (PDI) of the NPs were 242.9 nm and 0.074, respectively. The encapsulation efficiency and loading efficiency were 80.4 and 27.2%, respectively. Human embryonic kidney cells (HEK-293) were used for determining the cytotoxicity of papain-loaded PCL and PCL nanoparticles. The in vitro cell culture showed that nanoparticles are not toxic at low concentrations, while toxicity slightly increases at high concentrations. In silico studies, which were carried out with MD simulations and ADME analysis showed that the strong hydrogen bonds between the ligand and the papain provide stability and indicate the regions in which the interactions occur.     

Modeling of nonlinear ultrasound propagation in tissue from array transducers

A computationally efficient model capable of simulating finite-amplitude ultrasound beam propagation in water and in tissue from phased linear arrays and other transducers of arbitrary quasiplanar geometry is described. It is based on a second-order operator splitting approach [Tavakkoli et al., J. Acoust. Soc. Am. 104, 2061-2072 (1998)], with a fractional step-marching scheme, whereby the effects of diffraction, attenuation, and nonlinearity can be computed independently over incremental steps. This approach is an extension to that of Christopher and Parker [J. Acoust. Soc. Am. 90, 507-521; 90, 488-499 (1991)], wherein linear and nonlinear effects are propagated separately over incremental steps, and the computation of the diffractive substeps are based on an angular spectrum technique with a modified sampling scheme for accurate and efficient implementation of diffractive propagation from nonradially symmetric sources. Results of the model are compared with published data. Predicted field profiles for nonlinear propagation in tissue from realistic array transducers using the pulse inversion method are presented.