Nonlinear Reduction in Risk for Type 2 Diabetes by Magnesium Intake:An Updated Meta-Analysis of Prospective Cohort Studies

Observational studies between magnesium int- ake and risk of type 2 diabetes yielded inconsistent results. We conducted a system literature search of PubMed database through March 2015 for prospective cohort studies of magnesium intake and type 2 diabetes risk. Study-specific results were pooled in a random-effects model. Subgroup and sensitivity analysis were performed to assess the potential sources of heterogeneity and the robustness of the pooled estimation. Generalized least squares trend estimation was used to investigate the dose-response relationship. A total of 15 papers with 19 analyses were identified with 539,735 participants and 25,252 incident diabetes cases. Magnesium intake was associated with a significant lower risk of type 2 diabetes （RR： 0.77; 95% Ch 0.71-0.82） for the highest compared with lowest category. This association was not significantly modified by the pre-specified study characteristics. In the dose-response analysis, a magnesium intake increment of 100 mg/day was associated with a 16% reduction in type 2 diabetes risk （RR： 0.84; 95% Ch 0.80-0.88）. A nonlinear relationship existed between magnesium intake and type 2 diabetes （P-nonlinearity=0.003）. This meta-analysis further verified a protective effect of magnesium intake on type 2 diabetes in a nonlinear dose-response manner.

身体质量指数联合三酰甘油与高密度脂蛋白胆固醇比值预测2型糖尿病风险

目的利用体检指标中身体质量指数(BMI)和三酰甘油与高密度脂蛋白胆固醇比值(triglyceride to high-density lipoprotein cholesterol,TG/HDL-C)联合预测2型糖尿病(type 2 diabetes mellitus,T2DM)发病风险。方法基于瑞慈医疗集团2010—2016年体检人群的数据库,通过Cox比例风险模型观测不同BMI的人群基线TG/HDL-C对随访期间T2DM风险的影响,并分析TG/HDL-C与BMI是否存在交互作用。结果在随访期间,13685例研究对象中共有315例最终诊断为T2DM。以BMI=24 kg/m^(2)为界,将所有研究对象分成两个亚组。BMI<24 kg/m^(2)的亚组共有8238例研究对象,96例在随访结束时患T2DM,对潜在的混杂因素进行调整后,升高的TG/HDL-C导致T2DM事件的风险较高[HR 95%CI=1.47(1.23,1.74),P<0.001]。与最低五分位数(Q1)相比,升高的TG/HDL-C五分位数(Q2~Q5)T2DM发病率增加[HR 95%CI分别为:0.73(0.28,1.93)、2.75(1.29,5.85)、2.81(1.28,6.20)和3.91(1.51,10.11)]。BMI≥24 kg/m^(2)的亚组共有5447例研究对象,219例在随访结束时患T2DM。在两个亚组中,随着TG/HDL-C的增加,T2DM患病风险均呈非线性增加,这一变化在BMI<24 kg/m^(2)的亚组更为明显。绘制两个亚组TG/HDLC与3年和5年T2DM发病的ROC曲线,在BMI<24 kg/m^(2)的亚组中,ROC曲线下面积分别为0.72和0.77。在BMI≥24 kg/m^(2)的亚组中,ROC曲线下面积则分别为0.55和0.57。结论在BMI<24 kg/m^(2)的人群中,TG/HDL-C的升高提示T2DM患病风险显著增加。

中国股市在亚太地区的角色定位:基于油价冲击的风险溢出视角

国际油价波动对股票市场的影响研究一直备受关注,但对亚太和中国股市的风险溢出效应研究还有待拓展。鉴于冲击类型不同,本文将日度频率的油价波动分解为供给冲击、需求冲击和风险冲击,并基于基准回归模型与动态GARCH-Copula-CoVaR模型考察三类冲击对亚太和中国股市的影响与风险溢出。回归结果表明,需求冲击和风险冲击引起的油价上涨能显著提高两地区股市的平均收益,而供给冲击对此几乎无影响。从风险溢出视角看,中国股市在亚太地区主要扮演被动接受风险传染的角色:从需求冲击看,中国股市更易受到影响,呈现出风险净溢入的角色;但从风险冲击看,中国股市并不是联动亚太股市整体波动的源头,角色较为独立。结合上下行风险视角发现,两地区股市的上行风险总是显著大于下行风险,呈现出明显的非对称特征。但与亚太股市相比,中国股市风险值的波动更为剧烈。本文认为,辨析不同成因、来源的油价冲击,前瞻性预警亚太地区的油价波动风险,对于稳定我国股市、防范金融风险以及夯实我国在亚太地区的经济治理话语权具有重要意义。

钢框架-自复位SRC墙板结构的地震易损性分析

为了研究钢框架-型钢混凝土(SRC)自复位墙板结构的抗震性能,完成了钢框架-自复位SRC墙板子结构模型的低周往复加载试验,以研究其工作机理和失效机制。提出了基于宏观单元的非线性分析模型,并利用试验结果验证模型的准确性,研究结果表明该非线性分析模型能较好地模拟自复位SRC墙板的非线性行为。分别建立6层和15层钢框架-自复位SRC墙板结构的整体模型并进行增量动力时程分析(IDA)。以地面峰值加速度为地震动强度指标,以最大层间位移角为损伤指标,根据IDA结果得到不同极限状态下的易损性曲线,建立地震易损性矩阵,并对不同极限状态下结构的损伤风险进行了评估,研究表明:6层和15层钢框架-自复位SRC墙板结构在50年内的倒塌概率分别为0.138%和0.095%,远小于FEMA 750规定的1%,由此可见,钢框架-自复位SRC墙板结构抗震性能较好,并具备良好的抗地震倒塌能力。

降雨量与脑卒中入院的关联性:基于分布滞后非线性模型

背景脑卒中是严重危害人类健康的主要慢性病,降雨量与脑卒中发病的关系尚未明确。目的分析南昌市降雨量与脑卒中入院的关联性,为制定脑卒中综合防治策略和措施提供科学依据。方法收集江西省卫生健康委员会信息中心DRGs管理系统中南昌市2015—2019年脑卒中入院数据及全国城市空气质量实时发布平台大气污染物和南昌市气象基站数据,分析脑卒中入院患者、大气污染物及气象因素的基本特征。采用Spearman秩相关分析探究脑卒中入院例数与大气污染物和气象因素的相关性,利用分布滞后非线性模型分析降雨量与脑卒中入院的关系,并按照性别、年龄(<65岁和≥65岁)进行分层分析,lag代表滞后天数。结果2015—2019年南昌市脑卒中入院患者79523例,其中男性(49072例,61.71%)、年龄≥65岁(48092例,60.48%)的患者所占比例较大,冬(12月~次年2月)、春(3~5月)季脑卒中入院例数分别为20065例(25.23%)、20358例(25.60%)。降雨量与脑卒中入院呈非线性关系,并存在一定的滞后效应。降雨量对脑卒中入院的效应在lag1、lag2的RR值均为1.009,95%CI分别为1.000~1.019、1.001~1.016。分层分析显示:较高降雨量对男性脑卒中入院例数的主要效应为lag6,RR值为1.003;对女性脑卒中入院例数的主要效应为lag1和lag2,RR值分别为1.018(95%CI=1.004~1.031)、1.020(95%CI=1.009~1.031);对65岁以下缺血性脑卒中入院例数的主要效应为lag1(RR=1.016,95%CI=1.003~1.030)、lag2(RR=1.018,95%CI=1.007~1.029)。结论短期暴露于较高降雨量可增加脑卒中入院风险,女性、65岁以下人群对降雨暴露更为敏感,应加强对该人群的防护。

基于规则集成的可解释机器学习算法及应用

机器学习算法因其良好的预测性能已经取得了巨大的成功,但在对模型可解释性有着较高需求的领域,其适用性受到了限制。针对机器学习算法缺乏可解释性的缺点,基于规则集成思想提出一种新的可解释机器学习算法,称之为集成树惩罚逻辑规则回归,它能以较小的结构复杂度实现与集成树算法相当的预测性能,并且保留了逻辑回归的解释效果。首先,从随机森林、XGBoost等集成树中提取分枝,并将其转换为逻辑规则。其次,对规则集进行剪枝和去重处理,以得到精简的规则集。最后,将规则作为变量融入逻辑回归中,并以Lasso算法进行复杂度控制。以企业风险预警作为实例,与多种机器学习算法进行实验对比,结果表明此算法不仅能很好地继承集成树的违约判别能力,在各个分类指标上均超越了大多数机器学习算法,而且可以通过规则给出企业风险指标的阈值,便于企业进行风险管理。进一步地,根据此算法制作企业信用评分,验证了它的广泛适用性,得到的评分符合客观规律且具有区分度,然后通过三个公开数据集验证了模型预测性能的稳健性。

系统性金融风险对宏观经济运行的非线性影响研究

为了探究系统性金融风险对宏观经济的影响,该研究采用了CRITIC法和因子分析法,构建系统性金融风险指数和宏观经济韧性水平指数。通过MS-VAR模型探讨了不同区制状态下金融风险对宏观经济的非线性效应。研究发现,系统性金融风险对宏观经济存在时变影响,短期内对宏观经济造成负向冲击,但长期来看反而会提高经济韧性。研究结果还表明,系统性金融风险对宏观经济具有明显的时变特征,实时采取预防和控制措施,有序处置突出风险,有助于促进经济实现高质量发展。研究通过将经济韧性作为宏观经济抵御风险和冲击的评价指标,为宏观经济的发展提供了一个新的研究视角,同时为宏观经济抵御风险、促进经济的高质量发展提供了有力指导。

Decision-making model for risk management of cascade hydropower stations

In a medium-term electricity market,in order to reduce the risks of price and inflow uncertainties, the cascade hydropower stations may use the options contract with electricity supply companies. A profit-based model for risk management of cascade hydropower stations in the medium-term electricity market is presented. The objective function is profit maximization of cascade hydropower stations. In order to avoid the risks of price and inflow uncertainties, two different risk-aversion constraints: a minimum profit constraint and a minimum conditional value-at-risk, are introduced in the model. In addition, the model takes into account technology constraints of the generating units, which includes reservoir flow balance, reservoir capacity limits, water discharge constraints, etc. The model is formulated as a mixed integer nonlinear programming problem. Because the search space of the solution is very large, a genetic algorithm is used to deal with the problem.

Optimal Risk-Sensitive Filtering for System Stochastic of Second and Third Degree

The risk-sensitive filtering design problem with respect to the exponential mean-square cost criterion is con-sidered for stochastic Gaussian systems with polynomial of second and third degree drift terms and intensity parameters multiplying diffusion terms in the state and observations equations. The closed-form optimal fil-tering equations are obtained using quadratic value functions as solutions to the corresponding Focker- Plank-Kolmogorov equation. The performance of the obtained risk-sensitive filtering equations for stochastic polynomial systems of second and third degree is verified in a numerical example against the optimal po-lynomial filtering equations (and extended Kalman-Bucy for system polynomial of second degree), through comparing the exponential mean-square cost criterion values. The simulation results reveal strong advan-tages in favor of the designed risk-sensitive equations for some values of the intensity parameters.

中国上市金融机构风险传染研究——基于动态关联网络的视角

随着金融创新和金融科技的发展,金融机构作为金融体系的核心载体,相互间的关联网络日益紧密,跨机构的风险传染程度不断增加。从风险溢出视角,基于TENET模型构建风险溢出动态关联网络,深入研究金融机构间风险传染的方向和路径,结合2014~2021年我国金融机构间的相关数据,实证分析我国金融机构间非线性风险溢出网络结构和风险溢出水平。研究发现,我国金融机构系统性风险水平呈现周期性变化,非线性特征明显,其中证券部门的风险溢出强度最高,金融科技机构的系统性风险溢出逐渐增强,银行部门吸收了其他金融机构的大部分风险溢出,在维持金融系统稳定方面发挥了主要作用。当前,要进一步重视金融科技的监管应用,加强对系统性金融风险跨部门传染的监管,守住不发生系统性金融风险的底线。

基于尾部风险管理的期权动态套保策略研究

本文以满足一致性风险度量准则的CV aR为套保目标函数,采取从期货到期权的两步法估计期权动态最优套保比率,建立了基于尾部风险管理的期权动态套保模型,并以沪深300指数系列衍生品为样本开展了实证分析和稳健性检验。研究表明:(1)通过两步法建立的期权套保模型估计出的期权最优套保比率,既充分反映了衍生品市场与现货市场的动态相依关系,又有效克服了期权价格在不同到期阶段受标的资产价格之外因素的影响;(2)实证结果显示,不管沪深300E T F价格处于上涨周期还是下跌周期,期权动态套保效果均优于期货动态套保,充分体现了期权非线性特点在尾部风险管理中的优势;(3)策略选择上,在市场大幅波动背景下,多行权价期权组合套保效果优于单一行权价期权套保;(4)较宽的调仓阈值设置在市场涨跌幅较大时套保效果更显著。

Modeling optimal oil production paths under risk service contracts

Due to the rigorous fiscal terms and huge potential risk of risk service contracts,optimizing oil production paths is one of the main challenges in designing oilfield development plans.In this paper,an oil production path optimization model is developed to maximize economic benefits within constraints of technology factors and oil contracts.This analysis describes the effects of risk service contract terms on parameters of inputs and outputs and quantifies the relationships between production and production time,revenues,investment and costs.An oil service development and production project is illustrated in which the optimal production path under its own geological conditions and contract terms is calculated.The influences of oil price,service fees per barrel and operating costs on the optimal production have been examined by sensitivity analysis.The results show that the oil price has the largest impact on the optimal production,which is negatively related to oil price and positively related to service fees per barrel and operating costs.

基于机器学习技术的系统性金融风险监测预警

相对于传统的经济建模手段,机器学习方法具有良好的非线性建模性能等内在优点,从而显示出巨大的应用潜力。本文提出了一个基于机器学习技术的系统性金融风险监测预警体系,并从理论依据和实证检验双重角度给出了解释和验证。具体而言,从经济基本面、货币供给面、财政状况、证券和利率市场、价格指数、外汇和汇率市场、杠杆率及银行体系8个层面选取预测预警指标,并采用5种典型的机器学习模型及其集成模型对系统性金融风险进行预测,研究结果表明:(1)相对于传统的线性模型,善于捕捉非线性关系的机器学习模型,无论是在样本内还是在样本外均表现优异;(2)Lasso模型在向前一期预测时表现最好,SVM模型在向前多期预测时能力更强,集成模型则能兼顾样本内拟合效果和样本外预测能力,具有较强的稳健性;(3)PDP模型可以对特征的非线性性和重要性进行有效识别,从而有助于打开机器学习的黑箱;在所有预测预警变量中,汇率、货币供给量、市场利率以及工业品价格是影响系统性金融风险的关键因素,通过对这些重点变量的监测有助于对系统性金融风险进行早期防范。

Disaster Risk Reduction and Management: Recalling the Need for Paradigm Shift in Definition

The paper articulates the need for a paradigm shift in defining the basis of what constitutes a disaster. This new framework must be sensitive to the need for a more theoretical approach to inform disaster and hazard management. Central to this is that the resulting approach cannot be muddled in linear rational and procedural doctrines but appreciate the dynamics of complex nonlinearity of disaster events. By engaging in thought experiment and critical analysis of existing literature, the paper deconstructs the normative paradigm of defining disaster. The end is to inform disaster management and risk reduction intervention and mitigation programmes. The presented alternative approach is sensitive of the need to equally include;spatial, political, social, parameters. These are appreciated as being equally important as those dealing with ecological and economic. The resulting not only include other elements, but also expands the definition to the complex conditions inherent in the origin of phenomena to the interaction of the phenomena with multiple and complex socio-spatial and demographic dynamics, and then appreciating the complex results of this nonlinear interaction. The need for a more substantive definition of disaster underlines the pre-analysis that is necessary for implementation of mitigation and prevention strategies. That disasters are becoming more complex is synonymous with the complexities inherent in post-war development dynamics.

碳金融市场集成风险测度的新方法

文章选取核证减排量期货价格和欧元兑人民币汇率价格作为样本数据,基于非线性期望理论构建GE-Copula-VaR/CVaR模型,研究了碳金融市场各风险因子的不确定性与非线性相依性特征,并对碳金融市场集成风险进行动态测度。结果表明:研究期间的风险指标VaR与CVaR具有明显的时变特征,《巴黎协定》签署前的样本各风险因子波动存在更强的不确定性,碳金融市场风险较大;GE-Copula-CVaR模型在99%的置信水平下能通过Kupiec有效性检验,其准确性显著高于先进的非参数Copula-CVaR风险测度模型。

基于非线性模糊层次分析法的盾构施工风险评价研究

为解决突出影响因素对盾构施工风险评价结果造成影响的问题,构建基于组合赋权-非线性FAHP的盾构施工风险评价模型。基于工作分解结构法和风险分解结构法依次对盾构施工过程和风险来源分解,进而将二者结合构建盾构施工风险评价指标体系。通过乘法集成法将主、客观权重结合计算得到组合权重;结合专家打分法得到隶属度向量构造模糊关系矩阵;引入非线性算子综合计算分析组合权重和模糊关系矩阵;根据最大隶属度原则,确定盾构施工的最终风险等级。将构建的新模型应用到贵阳市轨道交通3号线1期工程桐木岭站—桃花寨站盾构施工进行风险实例分析,得出风险评价等级为“中度风险”,与现场实际得出的风险等级相符合,讨论了新模型与传统模糊层次分析法的区别,进一步验证了新模型的科学性、可靠性。

Tumor-Immune Interaction System with the Effect of Time Delay and Hyperglycemia on the Breast Cancer Cells

Breast cancer in women is a complicated and multifaceted disease. Studies have demonstrated that hyperglycemia is one of the most significant risk factors for breast cancer. Hyperglycemia is when the sugar level in human blood is too high, which means excess glucose. Glucose excess can encourage the growth, invasion, and migration of breast cancer cells at the cellular level. Though, the effects of glucose on the dynamics of breast cancer cells have been examined mathematically by a system of ordinary differential equations. However, the non-instantaneous biological occurrences leading to the secretion of immuno-suppressive cytokines by tumors to evade immune surveillance and the immune cells’ derivation of cytokines to attack the tumor cells are not yet discussed. Therefore, investigating the biological process involved in the dynamics of tumors, immune and normal cells with excessive glucose concentration is inviolable to determining the best procedure for controlling tumors’ uncontrollable growth. Time delay, denoted by τ, is used to describe the time tumor cells take to secrete immunosuppressive cytokines to evade immune surveillance and the time immune cells take to recognize and attack the tumor cells. We have studied the local stability analysis of the biological steady states in both delayed and non-delayed system. The Routh-Hurwitz stability criterion is used to analyze the dynamical equilibrium of the cells’ population. Hopf bifurcation was analyzed by using time delay s as a bifurcation parameter. The analytical results suggest an unstable scenario for a tumor-free equilibrium point as normal cells are bound to grow to their carrying capacity. The result predicts a stable system for coexisting equilibrium when the interaction is instantaneous (τ = 0). However, when τ > 0, the coexisting equilibrium point switches from stable to unstable. The numerical results not only validate all the analytical results but also show the case of possible situations when glucose concentration is varied, indicating that both tumor growth and immune system efficiency are highly affected by the level of glucose in the blood. This concluded that the delay in the secretion of cytokines by immune cells and derivation cytokines by the tumors helps to identify the possible chaotic situation under different glucose concentration and the extent to which such delay can have on restoration of the normal cells when glucose concentration is low.

媒体监督、流动性风险与失责行为的成本效应

根据失责行为曝光后的企业风险偏好特征,理论构建并实证检验了媒体监督、流动性风险对产品质量问题企业经营成本的经济效应。结果显示:媒体监督效力显著,且非线性影响企业经营成本;高风险偏好机制下,流动性不足的成本效应凸显,媒体监督效力约为低风险偏好机制的2倍;不同风险偏好机制间的阈值所有者权益占比为40.69%;曝光前、后,流动性风险都会增加企业经营成本,高风险偏好企业具有风险溢价效应。失责企业股权结构改革和重配应注重防范流动性风险。

人工神经网络在潜在地震危险区估计中的应用

利用环境剪应力较高的地区较容易发生中强地震这一原理,提出了将人工神经网络应用于估计潜在地震危险区的方法。由于该方法是通过人工神经网络技术提前1-2年预测某一地区未来环境应力值的变化来估计潜在地震危险区,因而可大大提高估计的准确性。该方法今后在多震地区的预报工作中值得一试。

基于偏度的多期证券投资组合模型研究

在考虑资产收益率分布中正的偏度水平前提下,以风险价值VaR为约束条件,并引入非线性交易费用、税收等市场摩擦因素,建立以累积偏度最大为目标函数的多期投资组合优化模型,用罚函数法和PSO算法结合求解此模型,并给出实证分析。考虑到在买卖资产风险时交易费用等对投资收益的影响,投资者应该在每一期都对其资产组合进行调整分析,确保在每一期的开始都建立起符合需要的最优资产组合,这对投资者的连续投资行为具有一定的指导意义。