Approximating Unknown Mappings: An Experimental Evaluation

Different methodologies have been introduced in recent years with the aim of approximating unknown functions. Basically, these methodologies are general frameworks for representing non-linear mappings from several input variables to several output variables. Research into this problem occurs in applied mathematics (multivariate function approximation), statistics (nonparametric multiple regression) and computer science (neural networks). However, since these methodologies have been proposed in different fields, most of the previous papers treat them in isolation, ignoring contributions in the other areas. In this paper we consider five well known approaches for function approximation. Specifically we target polynomial approximation, general additive models (Gam), local regression (Loess), multivariate additive regression splines (Mars) and artificial neural networks (Ann).Neural networks can be viewed as models of real systems, built by tuning parameters known as weights. In training the net, the problem is to find the weights that optimize its performance (i.e. to minimize the error over the training set). Although the most popular method for Ann training is back propagation, other optimization methods based on metaheuristics have recently been adapted to this problem, outperforming classical approaches. In this paper we propose a short term memory tabu search method, coupled with path relinking and BFGS (a gradient-based local NLP solver) to provide high quality solutions to this problem. The experimentation with 15 functions previously reported shows that a feed-forward neural network with one hidden layer, trained with our procedure, can compete with the best-known approximating methods. The experimental results also show the effectiveness of a new mechanism to avoid overfitting in neural network training.     

Thermodynamics of Mixtures Excess Volumes and Energies of Mixing of Cycloalkane Mixtures (Cyclopentane to Cyclooctane)

The volumes of mixing for the systems cyclopentane + cyclohexane, cyclopentane + cycloheptane, cyclohexane + cycloheptane (at 20°C), and cyclopentane + cyclooctane, cyclohexane + cyclooctane, cycloheptane + cyclooctane (at 30 °C) have been measured with a dilatometer. All systems show volume contractions, except the first. The energies of mixing are also given. They vary from 7 to 55 J/mol for equimolar solutions.     

Study of photocrosslinkable polysiloxanes bearing gem di-styrenyl groups: Synthesis and thermal properties

The synthesis of photocrosslinkable polysiloxanes containing gem di-oxaalkylene styrenyl groups and gem di-urethane-α-methyl styrenyl groups has been performed by copolycondensation of α,ω-dihydroxy polydimethyl siloxanes and dichlorosilanes bearing either cyclic acetal groups or Si-H groups (onto which the cyclic acetal groups are further added) and dichlorosilanes bearing alkyl groups. The introduction of styrenyl groups was then achieved by hydrolysis of the acetal groups into the corresponding alcohols followed by reaction with chloromethyl styrene or with 3-isopropenyl-α,α- dimethylbenzyl isocyanate.The structure of the different products synthesized was determined by IR, ¹H, ¹³C and ²⁹Si NMR spectroscopies. The thermal properties of the polysiloxanes bearing gem di-styrenyl groups have been studied at low and high temperatures.These products have been crosslinked under UV, in the presence of a cationic photoinitiator, and showed very good release paper properties.     

An expeditious hydroxyamidation of carboxylic acids

Capitalizing on in situ activation with the cyclic phosphonic anhydride PPAA (1), the conversion of carboxylic acids into hydroxamic acids has been reduced to an experimentally simple one-pot operation that addresses the issue of polyacylation without resorting to a large excess of hydroxylamine or to protection. Scope and selectivity were satisfactory with a wide range of substrates, including α,β-unsaturated acids and hydroxyacids.     

Pyrrolothieno[1,4]diazepines part III: Synthesis of amino,hydrazino and mercapto derivatives

Treatment of 2-(2-formyl-1H-pyrrol-1-yl)-3-thiophenecarbonitrile by sodium percarbonate afforded, in a one-pot sequence and on a multigram scale, 5,6-dihydro-6-hydroxy-4H-pyrrolo[1,2-a]thieno-[3,2-f][1,4]diazepin-4-one whose reactivity towards various nucleophiles like amines, hydrazines and thiols was evaluated.     

Synthesis,characterization and study of methyl 3-(2-oxo-2H-1,4-benzoxazin-3-yl) propanoate as new corrosion inhibitor for carbon steel in 1M H2SO4 solution

Research on Chemical Intermediates - A new benzoxazin derivative, namely methyl 3-(2-oxo-2H-1,4-benzoxazin-3-yl) propanoate (1,4-MBXP), was synthesized under mild conditions from 2-aminophenols and...     

3D-QSAR,molecular docking and ADMET studies of thioquinazolinone derivatives against breast cancer

Breast cancer is a deadly disease and the second largest cause of mortality on a worldwide platform. Despite the availability of several cancer treatments, life expectancies stay relatively poor. Consequently, the medicinal chemistry community prioritizes the quick discovery of novel anticancer drugs. In recent years, computational approaches have been widely used to accelerate the drug development process. In light of this, in the current work, we performed three-dimensional quantitative structure-activity relationships (3D-QSAR) and molecular docking analyses on thioquinazolinone derivatives with aromatase enzyme (PDB: 3S7S). External validation was used to validate the prediction capabilities of the generated model. The best CoMSIA (comparative molecular similarity indices analysis) model exhibited the significant values of $Q^2$, $R^{2}and$ $R_{pred}^2$. These findings suggested that the electrostatic, hydrophobic and hydrogen bond donor and acceptor fields have a significant effect on inhibition of breast cancer. Thus, a number of innovative potent aromatase inhibitors were designed and their biological activities were predicted based on the best model. Furthermore, molecular docking studies were carried out for the designed compounds against breast cancer. Additionally, ADMET proprieties were used to evaluate drug-likeness of these novel drug candidates. The most active compounds found by these computational studies could be helpful for synthesis and testing as prospective future anti-cancer treatments.     

An alternating direction method for general variational inequalities

In this paper, we present an alternating direction method for structured general variational inequalities. This method only needs functional values for given variables in the solution process and does not require the estimate of the co-coercive modulus. All the computing process are easily implemented and the global convergence is also presented under mild assumptions. Some preliminary computational results are given.