稀疏阈值的超分辨率图像重建

为了解决基于字典学习的超分辨重构算法耗时过长的问题,提出了基于稀疏阈值模型的图像超分辨率重建方法。首先,将联合字典理论与图像块稀疏阈值方法相结合,训练得到高、低分辨率过完备图像字典对。接着,通过稀疏阈值OMP算法对图像特征块进行稀疏表示。然后,通过高分辨率字典重构出初始的超分辨图像。最后,通过改进迭代反投影算法对初始的超分辨图像进行全局优化,从而进一步提高图像重构质量。实验结果表明,超分辨图像重构平均峰值信噪比(PSNR)为30.1 d B,平均结构自相似度(SSIM)为0.937 9,平均计算时间为10.2 s。有效提高了超分辨重构的速度,改善了重构高分辨图像的质量。     

An infrared small target detection algorithm based on high-speed local contrast method

Small-target detection in infrared imagery with a complex background is always an important task in remote sensing fields. It is important to improve the detection capabilities such as detection rate, false alarm rate, and speed. However, current algorithms usually improve one or two of the detection capabilities while sacrificing the other. In this letter, an Infrared (IR) small target detection algorithm with two layers inspired by Human Visual System (HVS) is proposed to balance those detection capabilities. The first layer uses high speed simplified local contrast method to select significant information. And the second layer uses machine learning classifier to separate targets from background clutters. Experimental results show the proposed algorithm pursue good performance in detection rate, false alarm rate and speed simultaneously.     

普适的基于能量的分块局域激发态聚类算法

在普适的基于能量的分块(GEBF)方法的框架下, 大体系的局域激发(LE)能可通过一系列活性子体系激发能的线性组合近似得到, 从而有效降低了计算的时间标度. 然而, 在体系的局域激发具有多个激发态的情形下, 如何有效识别所有活性子体系的激发特征并将其组合是一个挑战. 提出了一种基于局域激发态聚类的算法. 该方案基于空穴-电子分析和基于密度的聚类(DBSCAN)机器学习算法, 可以自动地聚合不同子体系中最相似的激发态并组合得到相应的局域激发态能量或激发能. 结合该算法改进的LE-GEBF方法在荧光分子衍生物、 荧光染料-水团簇及绿色荧光蛋白模型体系的计算中均获得了令人满意的结果. 该算法有望大大提升LE-GEBF方法在计算局域激发时的稳定性和准确性, 并可以有效处理吸收光谱具有多重峰的大体系.     

基于语义的软件服务故障自动诊断模型

模仿学习是机器人仿生机制研究的主要内容之一,即通过观察、理解、学习、模仿示教行为实现机器人的仿生特性。基于高斯过程分别表达采集离散示教信号所构成的示教轨迹和含有未知参数策略的模仿轨迹,构建模仿学习方法框架,将概率模型匹配引入到模仿学习中,以KL散度为代价函数比较两种轨迹的概率分布,运用梯度下降法寻求使KL散度最小的最优模仿控制策略,将策略应用于模仿机器人以完成与示教相同的模仿任务。以关节型机器人的机械臂摆动行为模仿为学习任务进行仿真,结果表明基于概率轨迹匹配的模仿学习方法能够实现机械臂摆动行为模仿,学习过程较传统方法简易且学习效果较好。     

纯电动汽车磷酸铁锂电池组的建模及优化

鉴于传统神经网络和支持向量机机理复杂、计算量大的缺陷,很难实时跟踪磷酸铁锂电池组复杂快速的内部反应,影响电池荷电状态的估算精度,提出应用一种简单、有效的极限学习机对一额定容量为100Ah、额定电压为72V的纯电动汽车磷酸铁锂电池组建模,并分别与BP神经网络、RBF神经网络、支持向量机进行对比。随后,以学习时间和泛化性能为优化目标,应用粒子群方法寻找最佳隐层节点个数。结果表明,基于极限学习机的磷酸铁锂电池组模型的学习时间、泛化性能优于BP神经网络、RBF神经网络、支持向量机;隐层节点优化后,模型的学习时间和泛化性能达到最优。     

糖厂澄清工段生产指标建模及操作参数优化

针对亚法糖厂澄清工段清汁色值和清汁残硫量难以在线测量的问题,提出了一种基于人工蜂群优化的在线极限学习机软测量方法。先用核主元分析法确定影响清汁质量的关键参数,建立基于在线极限学习机的软测量模型。同时利用人工蜂群算法对在线极限学习机的隐层参数进行寻优,优化所建模型。最后,使用带约束的粒子群对软测量模型进行优化求解,得到典型工况下的最优操作设定值,为后续工况操作提供参考依据。仿真结果表明,基于人工蜂群优化的在线极限学习机模型能够准确地预测清汁色值和残硫量,同时基于此模型优化的操作参数设定值能够达到期望的指标。     

Predicting Regioselectivity in Radical C−H Functionalization of Heterocycles through Machine Learning

Radical C−H bond functionalization provides a versatile approach for elaborating heterocyclic compounds. The synthetic design of this transformation relies heavily on the knowledge of regioselectivity, while a quantified and efficient regioselectivity prediction approach is still elusive. Herein, we report the feasibility of using a machine learning model to predict the transition state barrier from the computed properties of isolated reactants. This enables rapid and reliable regioselectivity prediction for radical C−H bond functionalization of heterocycles. The Random Forest model with physical organic features achieved 94.2 % site accuracy and 89.9 % selectivity accuracy in the out-of-sample test set. The prediction performance was further validated by comparing the machine learning results with additional substituents, heteroarene scaffolds and experimental observations. This work revealed that the combination of mechanism-based computational statistics and machine learning model can serve as a useful strategy for selectivity prediction of organic transformations.     

Taba: A Tool to Analyze the Binding Affinity

Evaluation of ligand-binding affinity using the atomic coordinates of a protein-ligand complex is a challenge from the computational point of view. The availability of crystallographic structures of complexes with binding affinity data opens the possibility to create machine-learning models targeted to a specific protein system. Here, we describe a new methodology that combines a mass-spring system approach with supervised machine-learning techniques to predict the binding affinity of protein-ligand complexes. The combination of these techniques allows exploring the scoring function space, generating a model targeted to a protein system of interest. The new model shows superior predictive performance when compared with classical scoring functions implemented in the programs Molegro Virtual Docker, AutoDock4, and AutoDock Vina. We implemented this methodology in a new program named Taba. Taba is implemented in Python and available to download under the GNU license at https://github.com/azevedolab/taba . © 2019 Wiley Periodicals, Inc.