改进的多层感知机在客户流失预测中的应用

针对在传统的客户流失预测数据预处理中，使用one-hot编码处理离散属性导致数据维度增加及数据过于稀疏的问题，提出了两种基于多层感知机的改进后的客户流失预测模型。其主要思想是分别使用堆叠自编码器和实体嵌入两种方法对多层感知机进行改进，通过将离散属性的高维编码数据向低维空间映射，有效地减少了one-hot编码产生的稀疏数据，增加了离散属性值之间的关联度。在对两份公开的数据集进行交叉验证后的实验结果表明，改进后的模型既有效地提高了预测的准确度，又维持了传统多层感知机模型在并行化计算方面的优势。     

Prioritization methodology of dangerous substances for water quality monitoring with scarce data

Monitoring and control of dangerous substances discharged into receiving waters have attracted more attention lately. Since it is not possible to analyze every single substance, a prioritization methodology is needed for the selection of those to be monitored. Existing well-developed models require significant amount of data for reliable outcomes. This paper presents a methodology to prioritize the dangerous substances having adverse effects on freshwaters in Turkey, where data are scarce. Such a methodology will also serve as a solid model for other countries with limited background data. The adopted methodology enabled the elimination of chemicals to generate a candidate list composed of 608 substances among more than 5000 substances. Further screening and prioritization were conducted using different assessment methods (i.e., Total Hazard Value, Total Impact Value, Combined Monitoring-based, and Modelling-based Priority Setting) to obtain a proposed Final Candidate Specific Pollutants List of 150 dangerous substances. The proposed Candidate National Pollutant List of Turkey was established by combining 45 priority pollutants of the European Union with a list of candidate specific pollutants. According to the outcomes of this study, monitoring and controlling of 195 dangerous substances in freshwaters are recommended. Further detailed studies should be conducted in order to observe the actual levels of these dangerous substances in freshwaters followed by a review of the monitoring list accordingly. Moreover, further revisions might be required in the proposed list due to some possible versatile conditions in terms of sampling points (i.e., change in the location of industries).     

悬吊式汽包锅炉膨胀中心计算模型改进及程序实现

该文对以前建立的膨胀中心计算模型做了改进 ,并以NG 410 /9.8 M型锅炉为例 ,讨论了其膨胀中心的设置以及导向载荷的计算方法及计算过程 ,重点论述了最小导向载荷、风载荷、地震载荷以及炉膛内压引起的不平衡载荷的计算。采用了竖直向下的计算模型。图 3表 4参 6     

高Q值光学微球腔的温度系数研究

光学微球谐振腔由于其具有超高的Q值及极小的模式体积等优点,在高灵敏度传感和光通信等方面得到了广泛的研究。测试了未封装和封装后微球腔谐振波长随温度的变化,实验结果表明随温度增大,谐振波长线性红移,且线性度高。二者温度系数不同,未封装时为25.6 pm/℃,封装后为4.4 pm/℃,主要原因为紫外胶的负热光系数所致。理论分析了紫外胶的热光效应,通过控制紫外胶厚度可以改变光在紫外胶中的比例,从而调节温度系数。当光在紫外胶中比例为0.1135时,温度系数变为0,可以抑制温度漂移,实现了温度补偿;该比例继续增大,温度灵敏度提高。低温漂、高灵敏度、微型化拓宽了回音壁模式(WGM)传感器的应用潜力。     

Mechanisms of Enhanced Catalysis in Enzyme–DNA Nanostructures Revealed through Molecular Simulations and Experimental Analysis

Understanding and controlling the molecular interactions between enzyme substrates and DNA nanostructures has important implications in the advancement of enzyme–DNA technologies as solutions in biocatalysis. Such hybrid nanostructures can be used to create enzyme systems with enhanced catalysis by controlling the local chemical and physical environments and the spatial organization of enzymes. Here we have used molecular simulations with corresponding experiments to describe a mechanism of enhanced catalysis due to locally increased substrate concentrations. With a series of DNA nanostructures conjugated to horseradish peroxidase, we show that binding interactions between substrates and the DNA structures can increase local substrate concentrations. Increased local substrate concentrations in HRP(DNA) nanostructures resulted in 2.9‐ and 2.4‐fold decreases in the apparent Michaelis constants of tetramethylbenzidine and 4‐aminophenol, substrates of HRP with tunable binding interactions to DNA nanostructures with dissociation constants in the micromolar range. Molecular simulations and kinetic analysis also revealed that increased local substrate concentrations enhanced the rates of substrate association. Identification of the mechanism of increased local concentration of substrates in close proximity to enzymes and their active sites adds to our understanding of nanostructured biocatalysis from which we can develop guidelines for enhancing catalysis in rationally designed systems.     

Pt nanoparticles synthesized with new surfactants: improvement in C1–C3 alcohol oxidation catalytic activity

Platinum electrocatalysts were prepared using PtCl₄ as a starting material and 1-decylamine, N,N-dimethyldecylamine, 1-dodecylamine, N,N-dimethyldodecylamine, 1-hexadecylamine, and 1-octadecylamine as surfactants. These surfactants were used for the first time in this synthesis to determine whether the primary and/or tertiary structure and/or chain length of the surfactants, affects the size and/or activity of the catalysts in C₁–C₃ alcohol electro-oxidation reactions. Electrochemical measurements (cyclic voltammetry and chronoamperometry) indicated that the highest electrocatalytic performance was observed for the Pt nanocatalysts that were stabilized by N,N-dimethyldecylamine, and this has a tertiary amine structure with a short chain length (R = C₁₀H₂₁). The high performance may be due to the high electrochemical surface area, Pt(0)/Pt(IV) ratio, %Pt utility, and roughness factor (R _f). X-ray photoelectron spectroscopy, X-ray diffraction, atomic force microscopy, and transmission electron microscopy were used to determine the parameters that affect the catalytic activities.     

Investigation of current‐voltage characteristics and current conduction mechanisms in composites of polyvinyl alcohol and bismuth oxide

Temperature dependent current‐voltage (I–V) measurements of Au/Polyvinyl Alcohol + Bi₂O₃/n‐Si structure were conducted between 100 and 350 K for investigating the temperature dependence of I–V characteristics and current conduction mechanisms in the structure. Series resistance of the structure is calculated using Ohm's law and Cheungs' method. Ideality factor (n) and zero‐bias barrier height (Φ_Bo) were obtained considering thermionic emission theory. From 100 to 350 K, n changed from 32.1 to 3.54, and Φ_Bo changed from 0.27 to 0.99 eV. Obtained temperature dependent values of n and Φ_Bo suggested that thermionic emission is not the dominant current conduction mechanism. Therefore, Ln(I)–Ln(V) curves of the studied structure were plotted for investigating current conduction mechanisms in the structure and current flow is explained considering space charge limited current. Moreover, density of interface states (D_it) in the structure were calculated and its temperature dependence was investigated such that D_it values are reduced to the order of ～10¹³ eV^?1 cm^?2 from ～10¹⁴ eV^?1 cm^?2 with increasing temperature. POLYM. ENG. SCI., 54:1811–1816, 2014. © 2013 Society of Plastics Engineers