Genetic training of network using chaos concept: application to QSAR studies of vibration modes of tetrahedral halides

The chaotic dynamical system is introduced in genetic algorithm to train ANN to formulate the CGANN algorithm. Logistic mapping as one of the most important chaotic dynamic mappings provides each new generation a high chance to hold GA's population diversity. This enhances the ability to overcome overfitting in training an ANN. The proposed CGANN has been used for QSAR studies to predict the tetrahedral modes (nu(1)(A1) and nu(2)(E)) of halides [MX(4)](epsilon). The frequencies predicted by QSAR were compared with those calculated by quantum chemistry methods including PM3, AM1, and MNDO/d. The possibility of improving the predictive ability of QSAR by including quantum chemistry parameters as feature variables has been investigated using tetrahedral tetrahalide examples.     

A chaotic approach to maintain the population diversity of genetic algorithm in network training

The concept of chaos being radically different from statistical randomness is introduced into chemometrics research. The chaotic system that is deterministic with underlying patterns and inherent ability in searching the space of interest has been employed to improve the performance of chemometric algorithms. In this paper, a chaotic mutation is introduced into the genetic algorithm (GA) which is used for artificial neural network (ANN) training. The chaotic algorithm is very efficient in maintaining the population diversity during the evolution process of GA. The proposed algorithm CGANN has been testified by prediction of vibrational frequencies of octahedral hexahalides from some selected molecular parameters.     

神经网络方法用于分辨3种化学物质

神经网络方法用于分辨３种化学物质冯伟,胡上序（浙江大学化工系,杭州,３１００２７）关键词模式分类,神经网络,模拟退火,遗传算法,传感器阵列传感器阵列技术是利用传感器阵列所提供的交叉敏和模式识别及微机处理技术,来提高传感器的选择性和传感器的测量精度［１...     

The lifetime of CFC substitutes studied by a network trained with chaotic mapping modified genetic algorithm and DFT calculations

The hydrohaloalkanes have attracted much attention as potential substitutes of chlorofluorocarbons (CFCs) that deplete the ozone layer and lead to great high global warming. Having a short atmospheric lifetime is very important for the potential substitutes that may also induce ozone depletion and yield high global warming gases to be put in use. Quantitative structure-activity relationship (QSAR) studies were presented for their lifetimes aided by the quantum chemistry parameters including net charges, Mulliken overlaps, E(HOMO) and E(LUMO) based on the density functional theory (DFT) at B3PW91 level, and the C-H bond dissociation energy based on AM1 calculations. Outstanding features of the logistic mapping, a simple chaotic system, especially the inherent ability to search the space of interest exhaustively have been utilized. The chaotic mapping aided genetic algorithm artificial neural network training scheme (CGANN) showed better performance than the conventional genetic algorithm ANN training when the structure of the data set was not favorable. The lifetimes of HFCs and HCs appeared to be greatly dependent on their energies of the highest occupied molecular orbitals. The perference of the RMSRE comparing to RMSE as objective function of ANN training was better for the samples of interest with relatively short lifetimes. C(2)H(6) and C(3)H(8) as potential green substitutes of CFCs present relatively short lifetimes.     

Application of the genetic algorithm joint with the Powell method to nonlinear least-squares fitting of powder EPR spectra

The application of the stochastic genetic algorithm (GA) in conjunction with the deterministic Powell search to analysis of the multicomponent powder EPR spectra based on computer simulation is described. This approach allows for automated extraction of the magnetic parameters and relative abundances of the component signals, from the nonlinear least-squares fitting of experimental spectra, with minimum outside intervention. The efficiency and robustness of GA alone and its hybrid variant with the Powell method was demonstrated using complex simulated and real EPR data sets. The unique capacity of the genetic algorithm for locating global minima, subsequently refined by the Powell method, allowed for successful fitting of the spectra. The influence of the population size, mutation, and crossover rates on the performance of GA was also investigated.     

The lifetime of CFC substitutes studied by a network trained with chaotic mapping modified genetic algorithm and DFT calculations

The hydrohaloalkanes have attracted much attention as potential substitutes of chlorofluorocarbons (CFCs) that deplete the ozone layer and lead to great high global warming. Having a short atmospheric lifetime is very important for the potential substitutes that may also induce ozone depletion and yield high global warming gases to be put in use. Quantitative structure–activity relationship (QSAR) studies were presented for their lifetimes aided by the quantum chemistry parameters including net charges, Mulliken overlaps, E _HOMO and E _LUMO based on the density functional theory (DFT) at B3PW91 level, and the C-H bond dissociation energy based on AM1 calculations. Outstanding features of the logistic mapping, a simple chaotic system, especially the inherent ability to search the space of interest exhaustively have been utilized. The chaotic mapping aided genetic algorithm artificial neural network training scheme (CGANN) showed better performance than the conventional genetic algorithm ANN training when the structure of the data set was not favorable. The lifetimes of HFCs and HCs appeared to be greatly dependent on their energies of the highest occupied molecular orbitals. The perference of the RMSRE comparing to RMSE as objective function of ANN training was better for the samples of interest with relatively short lifetimes. C₂H₆ and C₃H₈ as potential green substitutes of CFCs present relatively short lifetimes.     

Hybridized particle swarm algorithm for adaptive structure training of multilayer feed-forward neural network: QSAR studies of bioactivity of organic compounds

The multilayer feed-forward ANN is an important modeling technique used in QSAR studying. The training of ANN is usually carried out only to optimize the weights of the neural network and without paying attention to the network topology. Some other strategies used to train ANN are, first, to discover an optimum structure of the network, and then to find weights for an already defined structure. These methods tend to converge to local optima, and may also lead to overfitting. In this article, a hybridized particle swarm optimization (PSO) approach was applied to the neural network structure training (HPSONN). The continuous version of PSO was used for the weight training of ANN, and the modified discrete PSO was applied to find appropriate the network architecture. The network structure and connectivity are trained simultaneously. The two versions of PSO can jointly search the global optimal ANN architecture and weights. A new objective function is formulated to determine the appropriate network architecture and optimum value of the weights. The proposed HPSONN algorithm was used to predict carcinogenic potency of aromatic amines and biological activity of a series of distamycin and distamycin-like derivatives. The results were compared to those obtained by PSO and GA training in which the network architecture was kept fixed. The comparison demonstrated that the HPSONN is a useful tool for training ANN, which converges quickly towards the optimal position, and can avoid overfitting in some extent.     

Analyzing energy landscapes for folding model proteins

A new benchmark 20-bead HP model protein sequence (on a square lattice), which has 17 distinct but degenerate global minimum (GM) energy structures, has been studied using a genetic algorithm (GA). The relative probabilities of finding particular GM conformations are determined and related to the theoretical probability of generating these structures using a recoil growth constructor operator. It is found that for longer successful GA runs, the GM probability distribution is generally very different from the constructor probability, as other GA operators have had time to overcome any initial bias in the originally generated population of structures. Structural and metric relationships (e.g., Hamming distances) between the 17 distinct GM are investigated and used, in conjunction with data on the connectivities of the GM and the pathways that link them, to explain the GM probability distributions obtained by the GA. A comparison is made of searches where the sequence is defined in the normal (forward) and reverse directions. The ease of finding mirror image solutions are also compared. Finally, this approach is applied to rationalize the ease or difficulty of finding the GM for a number of standard benchmark HP sequences on the square lattice. It is shown that the relative probabilities of finding particular members of a set of degenerate global minima depend critically on the topography of the energy landscape in the vicinity of the GM, the connections and distances between the GM, and the nature of the operators used in the chosen search method.     

Global optimization of binary Lennard-Jones clusters using three perturbation operators

Global optimization of binary Lennard-Jones clusters is a challenging problem in computational chemistry. The difficulty lies in not only that there are enormous local minima on the potential energy surface but also that we must determine both the coordinate position and the atom type for each atom and thus have to deal with both continuous and combinatorial optimization. This paper presents a heuristic algorithm (denoted by 3OP) which makes extensive use of three perturbation operators. With these operators, the proposed 3OP algorithm can efficiently move from a poor local minimum to another better local minimum and detect the global minimum through a sequence of local minima with decreasing energy. The proposed 3OP algorithm has been evaluated on a set of 96 × 6 instances with up to 100 atoms. We have found most putative global minima listed in the Cambridge Cluster Database as well as discovering 12 new global minima missed in previous research.     

On the use of different potential energy functions in rare-gas cluster optimization by genetic algorithms: application to argon clusters

We study the effect of the potential energy function on the global minimum structures of argon clusters arising in the optimization performed by genetic algorithms (GAs). We propose a robust and efficient GA which allows for the calculation of all of the putative global minima of Ar(N) (N = 3-78) clusters modeled with four different potentials. Both energetic and structural properties of such minima are compared among each other and with those previously obtained for the Lennard-Jones function. In addition, the possibility of obtaining global minima of one potential through local optimization over the corresponding cluster geometry given by other potentials was associated with some structural parameters. The influence of the contribution from the three-body (Axilrod-Teller-Muto) triple-dipole potential (including or not a damping function to describe its correct behavior at smaller interatomic distances) has also been analyzed.     

Selection of quasi-optimal inputs in chemometrics modeling by artificial neural network analysis

Instrumentation spectra used for chemometrics analysis are often too unwieldy to model, as many of the inputs do not contain important information. Several mathematical methods are used for reducing the number of inputs to the significant ones only. Artificial neural networks (ANN) modeling suffers from difficulties in training models with a large number of inputs. However, using a non-random initial connection weight algorithm and local minima avoidance and escape techniques can overcome these difficulties. Once the ANN model is trained, the analysis of its connection weights can easily identify the more relevant inputs. Repeating the process of training the ANN model with the reduced input set and the selection of the more relevant inputs can proceed until a quasi-optimal, small, set of inputs is identified. Two examples are presented—finding the minimal set of wavelengths in benchmark diesel fuel NIR spectra, and in spectra generated in a recent work, modeling of “artificial nose” sensor array. In the last example, 1260 inputs were reduced to optimal sets of <10 inputs. Causal index calculation can analyze the influence of each of selected wavelengths on the predicted property. Some of the resulting minimal sets are not unique, depending on the ANN architecture used in the training. The accuracy of the resulting ANN models is usually better, and more robust, than the original large ANN model.     

On geometric-arithmetic index

The geometric-arithmetic (GA) index is a newly proposed graph invariant in mathematical chemistry. We give the lower and upper bounds for GA index of molecular graphs using the numbers of vertices and edges. We also determine the n-vertex molecular trees with the minimum, the second and the third minimum, as well as the second and the third maximum GA indices.     

遗传算法用于化学结构图的同态研究

提出一种采用整数串编码和基于节点基因交换方式的遗传算法,并应用于化学结构图的同态研究.遗传算法在一组随机生成的表示目标结构与查询结构节点间映射关系的整数串中进行逐步优化,直到找出与查询结构匹配的映射,从而实现化学结构图的同态匹配,并实现多重匹配.     

ANVAS: Artificial Neural Variables Adaptation System for descriptor selection

A new algorithm model-oriented for variable selection is presented in this study. It is based on the combination of genetic algorithms (GA) for hyperspace exploration, and counterpropagation artificial neural network (CP ANN) for deriving the fitness score. The proposed method performed very well on both well defined synthetic data sets and real academic data sets.     

基于旋转曲面变换PSO算法的神经网络用于胺类有机物毒性分类

神经网络模型能有效地模拟非线性的输入输出关系。本研究应用三层前馈网络对51种胺类有机物进行了结构-毒性关系的分类研究。常规的神经网络权值训练算法,例如误差反传算法,存在着收敛速度慢,容易陷入局部极值点等问题。因此提出旋转曲面变换粒子群优化算法,将被优化函数的局部极小点变换为全局最大点,同时不改变比局部极小点的值更小的区域的函数形状。此方法和粒子群优化相结合,能使待优化函数跳出局部极值点,提高训练神经网络权值的效率。实验结果显示,基于旋转曲面变换粒子群优化算法的神经网络,权值训练过程收敛速度较快,且自检误差和预报误差都较小,是一种有效的胺类有机物毒性分类方法。     

Artificial neural network data analysis for classification of soils based on their radionuclide content

The artificial neural network (ANN) data analysis method was used to recognize and classify soils of an unknown geographic origin. A total of 103 soil samples were differentiated into classes according to the regions in Serbia and Montenegro from which they were collected. Their radionuclide (²²⁶Ra, ²³⁸U, ²³⁵U, ⁴⁰K, ¹³⁴Cs, ¹³⁷Cs, ²³²Th, and ⁷Be) activities detected by gamma-ray spectrometry were then used as inputs to ANN. Five different training algorithms with different numbers of samples in training sets were tested and compared in order to find the one with the minimum root mean square error (RMSE). The best predictive power for the classification of soils from the fifteen regions was achieved using a network with seven hidden layer nodes and 2500 training epochs using the online back-propagation randomized training algorithm. With the optimized ANN, most soil samples not included in the ANN training data set were correctly classified at an average rate of 92%. The text was submitted by the authors in English.