基于电子病历的乳腺癌群组与治疗方案可视分析

乳腺癌是当前最常见的恶性肿瘤之一，其电子病历数据可用于挖掘隐含规律，对治疗与预后分析有重要意义。通过与乳腺科医生合作，选择合适的预测模型和可视化方法，搭建了一个基于电子病历的乳腺癌群组和治疗方案可视分析系统。首先，对具有高维属性的病人进行降维和聚类处理，形成病人群组，并采用南丁格尔图、词云和时间轴可视化方法，直观展示病人群组间特征的差异；然后，用支持向量机（support vector machine，SVM）模型预测治疗方案，用平行坐标、矩阵热力图和分类图分别展示属性相关性、训练后的特征权重和预测结果；最后，用真实案例验证了系统在群组分析、治疗方案及病人属性关联分析中的有效性，从而较好地帮助医生选择合适的治疗方案。     

两个基于2,5-二甲氧基对苯二甲酸和咪唑衍生物的超分子钴配位聚合物的结构及光催化反应

本文以咪唑衍生物为配体,通过水热合成法与钴离子制备出两个配位聚合物:{[Co(DTA)(1,4-DIB)(H₂O)]·H₂O}_n(1)和[Co(DTA)(1,3-BMIB)]_n(2)(1,4-DIB=1,4-二(1H-咪唑-1-基)苯; 1,3-BMIB=1,3-二(4-甲基-1H-咪唑-1-基)苯;H₂DTA=2,5-二甲氧基对苯二甲酸)。利用X射线单晶衍射、粉末衍射、热失重、元素分析、红外光谱以及固体紫外-可见光谱等对两个配合物进行了表征。结构分析证实配合物1和2是通过二维结构堆积成的三维超分子化合物。粉末衍射测试则显示两个配合物在水中有很好的稳定性。固体紫外-可见光谱显示两个配合物属半导体材料,对紫外-可见光有很强的吸收作用。在光催化实验中,配合物1和2可加快亚甲基蓝的降解速度。     

Mobility of water molecules in Li+, Na+ and Cs+ ionic forms of Nafion membrane studied by NMR

Correlation times and diffusion coefficients of water molecules were measured for the first time by ¹H spin relaxation and pulsed field gradient NMR in Li⁺, Na⁺ and Cs⁺ ionic forms of Nafion 117 membrane. Hydration numbers of Li⁺, Na⁺ and Cs⁺ cations were calculated. It was shown that at high humidity macroscopic transfer is controlled by the local translational motion of water molecules.     

基于快速神经网络架构搜索的鲁棒图像水印网络算法

为解决深度学习在图像水印算法中计算量大且模型冗余的问题，提高图像水印算法在抵抗噪声、旋转和剪裁等攻击时的鲁棒性，提出基于快速神经网络架构搜索（neural architecture search，NAS）的鲁棒图像水印网络算法。通过多项式分布学习快速神经网络架构搜索算法，在预设的搜索空间中搜索最优网络结构，进行图像水印的高效嵌入与鲁棒提取。首先，将子网络中线性连接的全卷积层设置为独立的神经单元结构，并参数化表示结构单元内节点的连接，预先设定结构单元内每个神经元操作的搜索空间；其次，在完成一个批次的数据集训练后，依据神经元操作中的被采样次数和平均损失函数值动态更新概率；最后，重新训练搜索完成的网络。水印网络模型的参数量较原始网络模型缩减了92%以上，大大缩短了模型训练时间。由于搜索得到的网络结构更为紧凑，本文算法具有较高的时间性能和较好的实验效果，在隐藏图像时，对空域信息的依赖比原始网络更少。对改进前后的2个网络进行了大量鲁棒性实验，对比发现，本文算法在CIFAR-10数据集上对抵抗椒盐噪声和旋转、移除像素行（列）等攻击优势显著；在ImageNet数据集上对抵抗椒盐高斯噪声、旋转、中值滤波、高斯滤波、JPEG压缩、裁剪等攻击优势显著，特别是对随机移除行（列）和椒盐噪声有较强的鲁棒性。     

紫色球杆菌视紫红质光谱特性的机器学习研究

近年来，机器学习等人工智能技术被应用于蛋白质工程，其在蛋白质结构、功能预测、催化活性等研究中具有独特优势。在未知蛋白质结构的情况下，将蛋白质序列和功能特性与机器学习相结合，基于序列-活性关系（innovative sequence-activity relationship，ISAR）算法，将蛋白质氨基酸序列数字化，用快速傅里叶变换（fast four transform，FFT）进行预处理，再进行偏最小二乘回归建模，可在数据集较少情况下拟合得到最佳模型。通过机器学习对紫色球杆菌视紫红质（gloeobacter violaceus rhodopsin，GR）的突变体蛋白质氨基酸序列与光谱最大吸收波长进行建模，获得了最佳模型。用最佳索引LEVM760106建模得到的确定系数R² 为0.944，均方误差E为11.64。用小波变换进行的预处理，其R² 虽也约为0.944，但E大于11.64，不及FFT进行的预处理。方法较好地解决了蛋白质序列与功能特性之间的数学建模问题，在蛋白质工程中可为预测更优的突变体提供支持。     

Vacuum‐driven fluid manipulation by a piezoelectric diaphragm micropump for microfluidic droplet generation with a rapid system response time

Recently, we developed a convenient microfluidic droplet generation device based on vacuum‐driven fluid manipulation with a piezoelectric diaphragm micropump. In the present study built on our previous work, we investigate the influence of settings applied to the piezoelectric pump, such as peak‐to‐peak drive voltage (V_p‐p) and wave frequency, on droplet generation characteristics. Stepwise adjustments to the drive voltage in ±10‐V_p‐p increments over the range of 200?250 V_p‐p during droplet creation revealed that the droplet generation rate could be reproducibly controlled at a specific drive voltage. The droplet generation rate switched within <0.5 s after the input of a new voltage. Although the droplet generation rate depended on the drive voltage, this setting had almost no influence on droplet size. The frequency over the selected range (50?60 Hz) did not markedly influence the droplet generation rate or droplet size. We show that the current fluid manipulation system can be conveniently used for both droplet generation and for rapid droplet reading, which is required in many microfluidic‐based applications.     

Impact of anisotropic slip on the stagnation-point flow past a stretching/shrinking surface of the Al_2O_3-Cu/H_2O hybrid nanofluid

The characteristics of heat transfer in the three-dimensional stagnationpoint flow past a stretching/shrinking surface of the Al_2O_3-Cu/H_2O hybrid nanofluid with anisotropic slip are investigated. The partial differential equations are converted into a system of ordinary differential equations by valid similarity transformations. The simplified mathematical model is solved computationally by the bvp4c approach in the MATLAB operating system. This solving method is capable of generating more than one solutions when suitable initial guesses are proposed. The results are proven to have dual solutions, which consequently lead to the application of a stability analysis that verifies the achievability of the first solution. The findings reveal infinite values of the dual solutions at several measured parameters causing the non-appearance of the turning points and the critical values. The skin friction increases with the addition of nanoparticles, while the escalation of the anisotropic slip effect causes a reduction in the heat transfer rate.     

Theoretical studies on the influence of metallic cations on ring opening of propylene oxide catalyzed by metal-salen complexes

We studied the ring opening of propylene oxide (PO) by salen-M coordinated OH⁻ group [M = Al(III), Sc(III), Cr(III), Mn(III), Fe(III), Co(II), Co(III), Ni(II), Cu(II), Zn(II), Ru(III) and Rh(III)]. The results show that the ring-opening energy barriers for M(II) complexes are much lower than those with M(III) complexes in the gas phase, and the barriers correlate linearly with the negative charges on the OH⁻ group and the Fukui function condensed on the OH⁻ group. The nucleophilicity ordering in the gas phase can be rationalized by the ratio of formal positive charges/radius of M cations. Solvent effect greatly increases the barriers of M(II) complexes but slightly changes the results of M(III) ones, making the barriers similar. Analysis indicates that the reaction heats are linearly proportional to the reverse reaction barriers. The relationships established here can be used to estimate the ring-opening barriers and to screen epoxide ring-opening catalysts.     

基于不确定理论的铁路货运需求预测

合理预测铁路货运需求是铁路管理部门建设、运营等决策基础。为应对铁路货运需求的复杂变化,基于Pearson相关性分析方法筛选出铁路货运需求的七个具有关键影响的因素,并结合不确定理论建立不确定多元线性回归模型,相应的铁路货运预测结果由传统单一值变成可能的需求区间范围,更加符合处于不确定环境下的铁路货运需求实际情况。选取国家统计局2004~2016年相关数据进实证研究,并与回归模型以及BP模型的预测结果对比分析,实验表明不确定多元线性回归的预测结果更加精确。