Les M. Sztandera

Soft Computing Approaches in Chemistry

Application of Evolutionary Algorithms to Combinatorial Library Design.- 1 Introduction.- 2 Overview of a Genetic Algorithm.- 3 De Novo Design.- 4 Combinatorial Synthesis.- 5 Combinatorial Library Design.- 6 Reactant Versus Product Based Library Design.- 7 Reactant-Based Combinatorial Library Design.- 8 Product-Based Combinatorial Library Design.- 9 Library-Based Designs.- 10 Designing Libraries on Multiple Properties.- 11 Conclusion.- References.- Clustering of Large Data Sets in the Life Sciences.- 1 Introduction.- 2 The Grouping Problem.- 3 Unsupervised Algorithms.- 4 Supervised Algorithms.- 5 Evaluation of Clustering Results.- 6 Interpretation of Clustering Results.- 7 Conclusion.- References.- Application of a Genetic Algorithm to the refinement of complex Mossbauer Spectra.- 1 Introduction.- 2 Theoretical.- 3 Experimental.- 4 Results.- 5 Discussion.- 6 Conclusions.- References.- Soft Computing, Molecular Orbital, and Functional Theory in the Design of Safe Chemicals.- 1 Introduction.- 2 Computational Methods.- 3 Neural Network Approach.- 4 Feed-Forward Neural Network Architecture.- 5 Azo Dye Database.- 6 Concluding Remarks.- Acknowledgement.- References.- Fuzzy Logic and Fuzzy Classification Techniques.- 1 Introduction.- 2 Fuzzy Sets.- 3 Case Studies of Fuzzy Classification Techniques.- 4 Conclusion.- References.- Further Reading.- Application of Artificial Neural Networks, Fuzzy Neural Networks, and Genetic Algorithms to Biochemical Engineering.- 1 Introduction.- 2 Application of Fuzzy Reasoning to the Temperature Control of the Sake Mashing Process.- 3 Conclusion.- Acknowledgements.- References.- Genetic Algorithms for the Geometry Optimization of Clusters and Nanoparticles.- 1 Introduction: Clusters and Cluster Modeling.- 2 Overview of Applications of GAs for Cluster Geometry Optimization.- 3 The Birmingham Cluster Genetic Algorithm Program.- 4 Applications of the Birmingham Cluster Genetic Algorithm Program.- 5 New Techniques.- 6 Concluding Remarks and Future Directions.- Acknowledgements.- References.- Real-Time Monitoring of Environmental Pollutants in the Workplace Using Neural Networks and FTIR Spectroscopy.- 1 Introduction.- 2 FTIR in the Detection of Pollutants.- 3 The Limitations of FTIR Spectra.- 4 Potential Advantages of Neural Network Analysis of IR Spectra.- 5 Application of the Neural Network to IR Spectral Recognition.- 6 Spectral Interpretation Using the Neural Network.- 7 Factors Influencing Network Performance.- 8 Comparison of Two and Three Layer Networks for Spectral Recognition.- 9 A Network for Analysis of the Spectrum of a Mixture of Two Compounds.- 10 Networks for Spectral Recognition and TLV Determination.- 11 Networks for Quantitative Spectral Analysis.- References.- Genetic Algorithm Evolution of Fuzzy Production Rules for the On-line Control of Phenol-Formaldehyde Resin Plants.- 1 Introduction.- 2 Resin Chemistry and Modelling.- 3 Simulation of Chemical Reactions.- 4 Model Comparison.- 5 Automated Control in Industrial Systems.- 6 Program Development.- 7 Comment.- References.- A Novel Approach to QSPR/QSAR Based on Neural Networks for Structures.- 1 Introduction.- 2 Recursive Neural Networks in QSPR/QSAR.- 3 Representational Issues.- 4 QSPR Analysis of Alkanes.- 5 QSAR Analysis of Benzodiazepines.- 6 Discussion.- 7 Conclusions.- References.- A Appendix.- Hybrid Modeling of Kinetics for Methanol Synthesis.- 1 Introduction.- 2 Neural Networks.- 3 Hybrid Modeling.- 4 Feature Selection.- 5 Modeling of Methanol Synthesis Kinetics.- 6 Conclusions.- A Appendix - Analytical Model of Methanol synthesis kinetics.- Acknowledgements.- References.- About the Editors.- List of Contributors.

Identification of the most significant comfort factors for textiles from processing mechanical, handfeel, fabric construction, and perceived tactile comfort data

Engineered fabrics are desired for military protective clothing applications. Such fabrics, exhibiting high tactile comfort, can be computationally designed. Through the use of an extensive database that contains handfeel, mechanical, construction, and tactile comfort data for fabrics, desired comfort can be predicted by measuring a limited number of properties. Output systems can be optimized to exhibit the highest level of comfort by engineering a fabric with specific properties. Using an extensive fabric database, we identify the most significant handfeel, mechanical, and construction properties influencing tactile fabric comfort. This is done through the use of regression analysis of handfeel, mechanical, construction fabric properties, and perceived tactile comfort, using B un-standardized coefficients and Beta standardized coefficients.

A neural network approach to prediction of glass transition temperature of polymers

Polymeric materials are finding increasing application in commercial optical communication systems. Taking advantage of techniques from the field of artificial intelligence, the goal of our research is to construct systems that can computationally design polymer formulations, including polymer optical fibers, with specified desirable consumer characteristics. Through the use of an extensive structure—property correlation database, properties of polymers can be predicted by an artificial network and the structure of novel polymers with desired properties can be optimized by a genetic algorithm. In this paper, we are focusing on one of the parameters, glass transition temperature (Tg) that influences a desired outcome in polymer optical fibers. Performance of such fibers can be optimized by engineering a polymer to exhibit a lower refractive index and Tg. This paper compares and discusses a neural network model and a linear model that have been developed to correlate Tg and repeating units of polymers. A neural network and multiple linear regression analysis were used in the study. A set of descriptors, chosen based on previous studies on the relations between Tg and polymer structure, were used to describe the structure of repeating units, individual bond energies, and intermolecular forces, especially hydrogen bonding, which is the strongest intermolecular force and exerts the greatest influence on Tg compared with other intermolecular interactions. A comprehensive neural network model with 28 descriptors was developed to predict Tg values of 6 randomly selected polymers from a database containing 71 polymers. The network was trained with the remaining 65 polymers and had a typical training root mean square error of 17 K (R2 = 0.95) and prediction average error of 17 K (R2 = 0.85). A linear regression model developed for comparison had an average error of 30 K (R2 = 0.88).

A neuro-fuzzy algorithm for diagnosis of coronary artery stenosis.

In this paper a method of fuzzy decision making applied to diagnosis of coronary artery stenosis is presented. The method uses a neural network approach for the diagnosis of stenosis in the three main coronary arteries (left anterior descending, right coronary artery, and circumflex). First, the knowledge base domain, 201Tl scintigram training data, is explained and the method of preprocessing the original heart images is given. Next, the method of dealing with the uncertainties present in the data using the fuzzy approach is outlined. Finally, the algorithm and the results are discussed and compared with other approaches.

Soft computing in textile sciences

Soft Computing for Softgoods Supply Chain Analysis and Decision Support.- Application of Fuzzy Set Theory in Mechanics of Composite Materials.- Soft Computing and Density Functional Theory in the Design of Safe Textile Chemicals.- Neural-Fuzzy Systems for Color Classifications in Textiles.- Agent-Based Modeling of the Textile/Apparel Marketplace.- Generating a Rule Set for the Fiber-to-Yarn Production Process by Means of an Efficiency-based Classifier System.- List of Contributors.- About the Editors.

Mutagenicity of aminoazo dyes and their reductive-cleavage metabolites: a QSAR/QPAR investigation

Abstract Quantitative structure–activity/property–activity relationships are developed that correlate the observed mutagenic behavior of 62 aminoazo derivatives and 12 of their reductive cleavage products with a variety of molecular descriptors calculated using quantum-chemical semiempirical methodology. Multilinear regression techniques using 8 descriptors are shown to account for more than 70% of the variation in the relative mutagenic activity of these compounds. Approaches using artificial neural networks in conjunction with fuzzy logic can account for about 95% of this variation using 8 descriptors.

Predicting Women’s Apparel Sales by Soft Computing

In this research, forecasting models were built based on both univariate and multivariate analysis. Models built on multivariate fuzzy logic analysis were better in comparison to those built on other models. The performance of the models was tested by comparing one of the goodness-of-fit statistics, R2, and also by comparing actual sales with the forecasted sales of different types of garments. Five months sales data (August-December 2001) was used as back cast data in our models and a forecast was made for one month of the year 2002. The performance of the models was tested by comparing one of the goodness-of-fit statistics, R2, and also by comparing actual sales with the forecasted sales. An R2 of 0.93 was obtained for multivariate analysis (0.75 for univariate analysis), which is significantly higher than those of 0.90 and 0.75 found for Single Seasonal Exponential Smoothing and Winters’ three parameter model, respectively. Yet another model, based on artificial neural network approach, gave an R2 averaging 0.82 for multivariate analysis and 0.92 for univariate analysis.

Analysis of tactile perceptions of textile materials using artificial intelligence techniques: Part 1: forward engineering

Purpose – The first of a two‐part series, this paper aims to discuss the design and development of an artificial intelligence‐based hybrid model to understand human perception of the tactile properties of textile materials and create an objective system to express those tactile perceptions in terms of measurable mechanical properties. Design/methodology/approach – A forward engineering system using the Model Free Algorithm approach of the Artificial Intelligence Technique to predict the tactile comfort score is presented. Findings – Human perception of tactile sensation is based on the weighted stimulus perceived by the human neural system. Originality/value – Contribution to intelligent textile and garment manufacture.

Monomethoxy-4-aminoazobenzenes: a computational study

Abstract The structural and electronic properties of the positional isomers of monomethoxy-4-aminoazobenzene (n-OMe-AAB) have been investigated using density functional theory with a basis set that includes polarization functions on all the atoms. These azo dyes are of interest because their carcinogenic activities depend dramatically on the position (n) of the methoxy group, e.g. 3-OMe-AAB is a potent hepatocarcinogen in the rat, whereas 2-OMe-AAB is a noncarcinogen. While it is generally believed that the various isomers of OMe-AAB require metabolic activation via N-hydroxylation prior to reaction with cellular macromolecules, we have shown that there are structural and electronic features present in these isomers that correlate with their carcinogenic behavior.

Analysis of tactile perceptions of textile materials using artificial intelligence techniques

Purpose – The second of a two‐part series, this paper aims to explain the design and development of a hybrid system for reverse engineering.Design/methodology/approach – A prediction engine to map the perception of tactile sensations using a neural network engine was developed. Since seventeen mechanical properties form the input ‐ and tactile compfort score is used as the output ‐ a direct reversal of the data set becomes impossible, hence, a hybrid approach was employed. The neural net is coupled with a genetic algorithm engine for the reversal process. The trained neural network acts as the objective function to evaluate the property set while the solution set is generated by Genetic Algorithm (GA) engine. Limitation of the GA and a means to overcome it is discussed. Application software based on the current research is also presented.Findings – Human perception of tactile sensations is non‐linear in terms of the mechanical properties of textile materials.Originality/value – The paper deals with revers...