基于特征融合的中文简历解析方法研究

针对基于规则和统计的传统中文简历解析方法效率低、成本高、泛化能力差的缺点，提出一种基于特征融合的中文简历解析方法，即级联Word2Vec生成的词向量和用BLSTM（Bidirectional Long Short-Term Memory）建模字序列生成的词向量，然后再结合BLSTM和CRF（Conditional Random Fields）对中文简历进行解析（BLSTM-CRF）。为了提高中文简历解析的效率，级联包含字序列信息的词向量和用Word2Vec生成的词向量，融合成一个新的词向量表示；再由BLSTM强大的学习能力融合词的上下文信息，输出所有可能标签序列的分值给CRF层；再由CRF引入标签之间约束关系求解最优序列。利用梯度下降算法训练神经网络，使用预先训练的词向量和Dropout优化神经网络，最终完成对中文简历的解析工作。实验结果表明，所提的特征融合方法优于传统的简历解析方法。     

Ballistic behavior of plain and reinforced concrete slabs under high velocity impact

This work presents a numerical simulation of ballistic penetration and high velocity impact behavior of plain and reinforced concrete slabs. In this paper, we focus on the comparison of the performance of the plain and reinforced concrete slabs of unconfined compressive strength 41 MPa under ballistic impact. The concrete slab has dimensions of 675 mm × 675 mm × 200 mm, and is meshed with 8-node hexahedron solid elements in the impact and outer zones. The ogive-nosed projectile is considered as rigid element that has a mass of 0.386 kg and a length of 152 mm. The applied velocities vary between 540 and 731 m/s. 6 mm of steel reinforcement bars were used in the reinforced concrete slabs. The constitutive material modeling of the concrete and steel reinforcement bars was performed using the Johnson-Holmquist-2 damage and the Johnson-Cook plasticity material models, respectively. The analysis was conducted using the commercial finite element package Abaqus/Explicit. Damage diameters and residual velocities obtained by the numerical model were compared with the experimental results and effect of steel reinforcement and projectile diameter were studies. The validation showed good agreement between the numerical and experimental results. The added steel reinforcements to the concrete samples were found efficient in terms of ballistic resistance comparing to the plain concrete sample.     

Effects of ethylene-octene copolymer (POE) on the brittle to ductile transition of high-density polyethylene/POE blends

Three kinds of ethylene-octene copolymers (POE) were melt-blended with high-density polyethylene (PE-HD) in different proportions. Detailed characterizations were conducted to analyze their structural differences of POE and its effects in toughening PE-HD. The higher molecular weight POE can improve the toughness of PE-HD. 60:40 PE-HD/POE is elongated to break up to 700% while impact strength is 84.7 kJ/m² at −30°C, which is 21-fold of PE-HD. In the brittle to ductile transition (BDT) during impact, the fracture mechanism changes from the crazing mode to the shear yield-plastic deformation mode. The BDT temperature decreases as the POE molecular weight and its content increase. The interface strength in tension is estimated to access their effects. The Boltzmann-type models were successfully extended to describe the typical S-shaped curves in BDT of notched impact strength vs POE content or temperature. The supplementary decay model is suggested for the attenuation in toughening. Transition map in impact is proposed to select the use range of composition (c ) and temperature (T ) for high toughness. The curves are converted into 3D graph of T -c -impact strength for illustrating their coupling-separate effects, and further into the contour map of impact strength in T -c space for finding their partial equivalence.     

Detection of correlation characteristics between financial time series based on multi-resolution analysis

Interactions between financial time series are complex and changeable in both time and frequency domains. To reveal the evolution characteristics of the time-varying relations between bivariate time series from a multi-resolution perspective, this study introduces an approach combining wavelet analysis and complex networks. In addition, to reduce the influence the phase lag between the time series has on the correlations, we propose dynamic time-warping (DTW) correlation coefficients to reflect the correlation degree between bivariate time series. Unlike previous studies that symbolized the time series only based on the correlation strength, the second-level symbol is set according to the correlation length during the coarse-graining process. This study presents a novel method to analyze bivariate time series and provides more information for investors and decision makers when investing in the stock market. We choose the closing prices of two stocks in China’s market as the sample and explore the evolutionary behavior of correlation modes from different resolutions. Furthermore, we perform experiments to discover the critical correlation modes between the bull market and the bear market on the high-resolution scale, the clustering effect during the financial crisis on the middle-resolution scale, and the potential pseudo period on the low-resolution scale. The experimental results exactly match reality, which provides powerful evidence to prove that our method is effective in financial time series analysis.     

腹腔手术后重症患者心肌损伤的危险因素

目的探讨腹腔手术后重症患者心肌损伤的发生情况及可能的危险因素。方法回顾性分析北京大学人民医院2017年1月至2019年1月腹腔手术后重症患者的一般临床资料及心肌损伤情况,收集并观察基础病史、术中(手术时间、是否急诊手术、术中出血>800 ml和术中低血压等)及术后指标(改良氧合指数、血乳酸、急性肾损伤和术后24 h内使用升压药情况等)。根据术后是否发生心肌损伤,将患者分为心肌损伤组和非心肌损伤组,采用Logistic回归分析腹腔手术后重症患者心肌损伤的危险因素。结果在纳入的803例腹腔手术后重症患者中,心肌损伤发生率为17. 2%(138/803),而急性心肌梗死发生率仅为0. 9%(7/803)。单因素分析显示,慢性肾功能不全病史、手术时间、急诊手术、术中低血压、术后24 h内使用升压药、高APACHEⅡ评分及术后即刻急性肾损伤与术后重症患者心肌损伤相关(P<0. 05)。多因素回归分析显示,急诊手术(OR=3. 14,95%CI:1. 76～5. 60,P<0. 001)、术后24 h内使用升压药(OR=2. 26,95%CI:1. 23～4. 15,P=0. 008)、APACHEⅡ评分(OR=1. 05,95%CI:1. 01～1. 09,P=0. 008)和术后急性肾损伤(OR=3. 18,95%CI:1. 78～5. 69,P <0. 001)与腹腔手术后重症患者发生心肌损伤独立相关。结论重症患者腹腔手术后心肌损伤发生率高,急诊手术、术后24 h内使用升压药、高APACHEⅡ评分和术后急性肾损伤是导致腹腔手术后重症患者发生心肌损伤的独立危险因素。     

Pull out of FRP reinforcement from masonry pillars: Experimental and numerical results

In this paper, debonding phenomena between carbon fiber reinforced polymer (CFRP) strips and masonry support were investigated on the basis of single-lap shear tests, considering different dimensions of the bond length. To capture the post-peak response of the CFRP–masonry joint, the slip between the support and the reinforcement strip was controlled using a clip gauge positioned at the end of the reinforcement. The tests were simulated by means of a finite element model able to capture the post-peak snap-back behavior due to the failure process. The numerical model is based on zero-thickness interface elements and on a proper non-linear cohesive law. The comparison between experimental and numerical results was performed in terms of overall response, measured by both the machine stroke and the clip gauge positioned at the free end of the reinforcement. The cases of effective bond length greater and lesser than the minimum anchorage length, suggested by the CNR Italian recommendation, were considered.     

Application of Vis–NIR hyperspectral imaging in classification between fresh and frozen-thawed pork Longissimus Dorsi muscles

Fresh and frozen-thawed (F-T) pork meats were classified by Vis–NIR hyperspectral imaging. Eight optimal wavelengths (624, 673, 460, 588, 583, 448, 552 and 609 nm) were selected by successive projections algorithm (SPA). The first three principal components (PCs) obtained by principal component analysis (PCA) accounted for over 99.98% of variance. Gray-level-gradient co-occurrence matrix (GLGCM) was applied to extract 45 textural features from the PC images. The correct classification rate (CCR) was employed to evaluate the performance of the partial least squares-discriminate analysis (PLS-DA) models, by using (A) the reflected spectra at full wavelengths and (B) those at the optimal wavelengths, (C) the extracted textures based on the PC images, and (D) the fused variables combining spectra at the optimal wavelengths and textures. The results showed that the best CCR of 97.73% was achieved by applying (D), confirming the high potential of textures for fresh and F-T meat discrimination.     

因素空间理论下基点分类算法研究

目前，基于因素空间理论的背景基提取算法计算过程复杂，初始化必须依赖各因素极值，基点数量提取冗余等原因，未能在应用中取得很好效果。为此，结合内点判别法和知识可继承、可扩展的思想，提出一种计算简单、初始化独立、基点数量小的改进的背景基提取算法。然后，利用改进的背景基提取算法构造出一种全新的数据分类算法-基点分类算法，基点分类算法以提取每一类样本的背景基为预测模型，再通过新定义的λ-背景基，优化预测模型。数值实验表明:基点分类算法原理简单、构造难度小、分类模型泛化能力强，预测能力准确率高，同时严格的模型限定区域又能为识别新类别提供新方法。