阿尔茨海默病诊断的自适应BP神经网络模型

阿尔茨海默病(AD)和轻度认知功能损伤(MCI)具有患者多、诊断难的特点,改进BP神经网络,提出自适应BP神经网络(ABP)进行100次AD和MCI诊断模拟,ABP神经网络的诊断正确率显著高于BP和RBF神经网络.采用留一法将101例正常人、200例MCI和90例AD患者的样本分为训练集和检测集,用ABP神经网络对其进行诊断模拟,总正确率达到73.91%.     

An improved ensemble learning machine for biological activity prediction of tyrosine kinase inhibitors

Boosting is one of the most important strategies in ensemble learning because of its ability to improve the stability and performance of weak learners. It is nonparametric, multivariate, fast and interpretable but is not robust against outliers. To enhance its prediction accuracy as well as immunize it against outliers, a modified version of a boosting algorithm (AdaBoost R2) was developed and called AdaBoost R3. In the sampling step, extremum samples were added to the boosting set. In the robustness step, a modified Huber loss function was applied to overcome the outlier problem. In the output step, a deterministic threshold was used to guarantee that bad predictions do not participate in the final output. The performance of the modified algorithm was investigated with two anticancer data sets of tyrosine kinase inhibitors, and the mechanism of inhibition was studied using the relative weighted variable importance procedure. Investigating the effect of base learner's strength reveals that boosting is only successful using the classification and regression tree method (a weak to moderate learner) and does not have a significant effect using the radial basis functions partial least square method (a strong base learners). Copyright © 2015 John Wiley & Sons, Ltd.     

Search for a Steric Penultimate Effect: Copolymerization of 2,3,4-Trimethyl-3-pentyl Methacrylate with Styrene

The hindered monomer, 2,3,4-trimethyl-3-pentyl methacrylate (I), was synthesized for penultimate effect studies. Since it readily homopoiymerized (km¹¹¹≠ 0) and readily copolymerized with styrene, copolymerizations of I with styrene were carried out at 60°C in benzene with AIBN as initiator. The conversion to copolymer and the copolymer composition were determined by using GLC techniques. Composition-conversion data was analyzed by performing a computerized nonlinear least-squares fitting to the integrated form of the penultimate model equation. The experimental design included the use of optimized M¹°/M²° ratios. The penultimate reactivity ratios calculated from these data were r¹′ = 0.23, r¹′= 0.59, r² = 0.59, r²′ = 1.34. Thus, when I is the penultimate unit, a terminal styryl radical prefers to add styrene, whereas when styrene is the penultimate unit, terminal styryl radicals prefer to add I. These results constitute the best evidence for a steric penultimate effect yet available in the literature from composition-conversion studies. However, the case is not yet proved. Further studies to strengthen this conclusion are proposed.     

Switching adaptive controllers to control fractional‐order complex systems with unknown structure and input nonlinearities

This article investigates the chaos control problem for the fractional‐order chaotic systems containing unknown structure and input nonlinearities. Two types of nonlinearity in the control input are considered. In the first case, a general continuous nonlinearity input is supposed in the controller, and in the second case, the unknown dead‐zone input is included. In each case, a proper switching adaptive controller is introduced to stabilize the fractional‐order chaotic system in the presence of unknown parameters and uncertainties. The control methods are designed based on the boundedness property of the chaotic system's states, where, in the proposed methods the nonlinear/linear dynamic terms of the fractional‐order chaotic systems are assumed to be fully unknown. The analytical results of the mentioned techniques are proved by the stability analysis theorem of fractional‐order systems and the adaptive control method. In addition, as an application of the proposed methods, single input adaptive controllers are adopted for control of a class of three‐dimensional nonlinear fractional‐order chaotic systems. And finally, some numerical examples illustrate the correctness of the analytical results. © 2014 Wiley Periodicals, Inc. Complexity 21: 211–223, 2015     

基于自适应模糊控制的节能装置研究

针对凹版印刷干燥设备,设计节能装置,该装置利用压缩机换热原理,对余热进行再循环利用。设计自适应模糊PID节能控制系统,完成对压缩机,调功器,变频器的实时控制。利用组态软件完成系统界面设计,人机交互简洁,操作方便。控制算法采用自适应模糊PID控制,依靠数据信息来调整模糊逻辑系统的参数,通过在线计算得到最佳控制器参数,抑制不确定性对系统的影响。将节能装置应用到凹版印刷机,实验数据显示,系统耗能减少,效率得到提高,达到节能目的,证明了节能系统的有效性,具有较强的应用价值。     

二冷配水优化建模与混合自适应PSO算法求解

针对连铸二冷区生产环境复杂且存在着大量水雾干扰的情况,建立了连铸水量优化模型并提出了一种混合的自适应粒子群算法来求解连铸二冷水优化问题。依据冶金过程中的工艺要求建立了二冷水量优化模型,并在经典的PSO算法基础上提出了适合该问题求解了混合自适应PSO算法。由于连铸过程存在着偏微分方程约束,传统的优化方法容易陷入局部最优解,不能达到很好的动态优化效果。研究了粒子群算法,基于种群的多样性,不断的自适应的更新粒子群算法中参数,将禁忌搜索的方法和传统的粒子群算法结合,增强了算法的局部搜索能力和全局寻找全局最优的能力。将该算法应用到连铸二冷水动态优化中,实验结果表面该算法能够快速有效的求解该优化问题。该方法用于连铸二冷水优化是可行的、有效的。     

A posterior error estimates for the nonlinear grating problem with transparent boundary condition

The nonlinear grating problem is modeled by Maxwell's equations with transparent boundary conditions. The nonlocal boundary operators are truncated by taking sufficiently many terms in the corresponding expansions. A finite element method with the truncation operators is developed for solving the nonlinear grating problem. The two posterior error estimates are established. The a posterior error estimate consists of two parts: finite element discretization error and the truncation error of the nonlocal boundary operators. In particular, the truncation error caused by truncation operations is exponentially decayed when the parameter N is increased. Numerical experiment is included to illustrate the efficiency of the method. © 2014 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 31: 1101–1118, 2015     

Accurate prediction of relative binding affinities of a series of HIV-1 protease inhibitors using semi-empirical quantum mechanical charge

A major challenge in computer-aided drug design is the accurate estimation of ligand binding affinity. Here, a new approach that combines the adaptive steered molecular dynamics (ASMD) and partial atomic charges calculated by semi-empirical quantum mechanics (SQMPC), namely ASMD-SQMPC, is suggested to predict the ligand binding affinities, with 24 HIV-1 protease inhibitors as testing examples. In the ASMD-SQMPC, the relative binding free energy (ΔG) is reflected by the average maximum potential of mean force (<PMF>_max) between bound and unbound states. The correlation coefficient (R²) between the <PMF>_max and experimentally determined ΔG is 0.86, showing a significant improvement compared with the conventional ASMD (R² = 0.52). Therefore, this study provides an efficient approach to predict the relative ΔG and reveals the significance of precise partial atomic charges in the theoretical simulations.     

Moving boundary truncated grid method: Multidimensional quantum dynamics

The moving boundary truncated grid (TG) method is used to study wave packet dynamics of multidimensional quantum systems. As time evolves, appropriate Eulerian grid points required for propagating a wave packet are activated and deactivated with no advance information about the dynamics. This method is applied to the Henon-Heiles potential and wave packet barrier scattering in two, three, and four dimensions. Computational results demonstrate that the TG method not only leads to a great reduction in the number of grid points needed to perform accurate calculations but also is computationally more efficient than the full grid calculations.