投资组合优化问题中的输入参数大多是由历史数据估计而来,估计的不确定性可能对Markowitz投资组合模型产生巨大的影响.近期,一个联合估计与鲁棒性的优化框架(joint estimation and robustness optimization,JERO)被提出,通过结合参数估...
详细信息
投资组合优化问题中的输入参数大多是由历史数据估计而来,估计的不确定性可能对Markowitz投资组合模型产生巨大的影响.近期,一个联合估计与鲁棒性的优化框架(joint estimation and robustness optimization,JERO)被提出,通过结合参数估计和优化问题以减弱估计不确定性对优化问题的影响.JERO框架被应用到投资组合优化领域(JERO with the mean return and the risk(variance)constraints,JERO-MV),同时考量了投资组合模型中有价值的两个度量:投资组合的回报和风险.但该模型可能会导致投资组合过分集中于某几个资产,这将增加投资风险和成本.本文在JERO-MV模型的基础上增加分散化约束,并给出该模型的可行性条件.本文在真实数据集上进行了大量的数值实验,并与JERO-MV模型进行对比.在大多数情形下,我们的模型都有更好的样本外表现.
图结构学习(Graph Structure Learning, GSL)通过优化图结构,增强图的表示能力和性能。GSL能够更好地捕捉图数据中节点之间的关系,从而促进信息的有效传播。图结构优化在商品链接预测中的应用研究旨在通过改进商品间关系的图结构,提高预测精度与推荐效果。在电商平台中,商品间的复杂关系往往通过图结构表示,其中节点代表商品,边代表商品间的关联或共同特征。通过优化图的构建和学习方法,能够更准确地捕捉商品之间的潜在联系,从而提升链接预测的准确性和推荐质量。优化后的图结构可以帮助算法更好地处理大规模商品数据,增强模型的泛化能力,进而提升电商平台的个性化推荐系统,增加用户购买的可能性,并促进销售增长。本文提出了一种新的稀疏正则化与图结构学习模型搜索方法(SGSL)。通过引入边缘修剪的正则化项等技术,SGSL能够在节点不变分类任务中显著提高性能,同时减少在节点变化任务中搜索到错误边的风险。实验表明,SGSL能有效增强图神经网络模型的性能。Graph Structure Learning (GSL) enhances the representational capacity and performance of graphs by optimizing their structure. GSL better captures the relationships between nodes in graph data, which facilitates more effective information propagation. The application of graph structure optimization in product link prediction aims to improve prediction accuracy and recommendation performance by refining the graph structure that represents the relationships between products. In e-commerce platforms, the complex relationships between products are often represented through graph structures, where nodes represent products and edges represent associations or shared features. By optimizing graph construction and learning methods, the underlying relationships between products can be more accurately captured, thereby improving link prediction accuracy and recommendation quality. The optimized graph structure helps algorithms better handle large-scale product data, enhancing the model’s generalization ability, which in turn improves personalized recommendation systems, increases the likelihood of user purchases, and drives sales growth. This paper introduces a novel Sparse Regularization and Graph Structure Learning Model Search method (SGSL). By incorporating techniques such as edge pruning regularization, SGSL significantly improves performance in node-invariant classification tasks while reducing the risk of selecting incorrect edges in node-variant tasks. Experimental results show that SGSL effectively enhances the performance of graph neural network mo
本文探讨了双碳政策实施背景下,中国能源消耗结构转型与国际贸易的相互影响。随着全球对可持续发展的重视,中国正加速向可再生能源转型,逐步降低对化石燃料的依赖。这一转型不仅改善了国内能源安全,也重塑了国际能源贸易格局。文章基于线性回归模型预测了中国的碳排放量,并分析了近十年来的能源消耗模式和结构变化。同时,通过相关性分析揭示了能源消耗与国际贸易之间的内在联系,指出两者在全球绿色经济中的互动关系。中国的转型不仅有助于改善环境状况,还为国际市场带来了新的机遇与挑战。随着可再生能源比重的提升,中国在全球能源供应链中的地位也在发生变化。这一变化促使中国企业在技术创新和国际合作方面不断提升,从而推动国际贸易向更高标准和更环保的方向发展。总的来说,本文为理解中国在全球能源转型中的角色提供了重要洞见,强调了可持续发展与国际贸易之间的密切联系。This paper examines the relationship between the structural transformation of China’s energy consumption and international trade, with a particular focus on the implementation of peak carbon and carbon-neutral policy policies. In light of the global emphasis on sustainable development, China is accelerating its transition to renewable energy sources and gradually reducing its dependence on fossil fuels. This transition has the additional benefit of enhancing domestic energy security while simultaneously influencing the configuration of international energy trade patterns. The article employs a linear regression model to predict China’s carbon emissions and conducts an in-depth analysis of the shifts in energy consumption patterns and structures that have occurred over the past decade. Concurrently, correlation analyses elucidate the intrinsic interconnections between energy consumption and international trade, underscoring the interplay between the two in the global green economy. China’s transition not only has positive implications for environmental conditions, but also presents new opportunities and challenges for the international market. As the proportion of renewable energy sources increases, China’s role in the global energy supply chain is evolving. This shift has prompted Chinese companies to enhance their technological innovation and international collaboration, thereby propelling international trade towards higher standards and a more environmentally conscious approach. Overall, this paper offers valuable insights into China’s role in the global energy transition, underscoring the intrinsic link
本文将比例风险模型应用于用户下单时长的研究,揭示影响下单时长的变量及其对用户下单决策的影响。利用比例风险模型识别影响用户下单时长的变量并对其进行参数估计,如商品价格、用户情况、促销信息、用户个体差异等。并通过实例分析评估各因素的影响程度,量化各个因素对用户下单时长的影响,并分析这些因素如何提高或延缓用户的下单决策。基于模型结果,提出优化建议,帮助电商平台改善用户体验和转化率。This paper applies the proportional hazards model to the study of the time it takes for users to place orders, revealing the variables that influence the time and their impact on users’ ordering decisions. The proportional hazards model is used to identify the variables that influence the time it takes for users to place orders and to estimate their parameters, such as product prices, user conditions, promotional information, and individual differences of users. The paper also analyzes the degree of influence of each factor through an example and quantifies the impact of each factor on the time it takes for users to place orders. It further analyzes how these factors can enhance or delay users’ ordering decisions. Based on the results of the model, the paper proposes optimization suggestions to help e-commerce platforms improve user experience and conversion rates.
水果作为我国种植业中第四大农作物类别,在国民经济中占据着十分重要的地位,水果的价格波动不仅影响居民生活成本,也影响果农的收入。因此,预判水果价格的走势十分重要。文中选取2021年第22周至2024年第38周,共计174周的红富士苹果、巨峰葡萄和麒麟西瓜这三种水果的周度批发价格数据,建立VMD-SSA-LSTM模型,并与LSTM、VMD-LSTM模型的水果价格预测效果进行对比,实验发现,VMD-SSA-LSTM模型的预测值最接近真实值,且其RMSE、MAE、MAPE的数值均小于LSTM、VMD-LSTM模型的数值,表明VMD-SSA-LSTM模型的预测效果优于LSTM、VMD-LSTM模型。Fruits are the fourth largest crop category in China’s agricultural sector and play a vital role in the national economy. The fluctuations in fruit prices not only affect the cost of living for residents but also affect the income of fruit farmers. Therefore, it is crucial to forecast the trend of fruit prices. In this study, we selected the weekly wholesale prices of Red Fuji apples, Jufeng grapes, and Qilin watermelons from week 22 of 2021 to week 38 of 2024, a total of 174 weeks, and built the VMD-SSA-LSTM model to predict the prices of these three fruits. We also compared the fruit price prediction effects of the VMD-SSA-LSTM model with those of the LSTM and VMD-LSTM models. The experimental results show that the predicted values of the VMD-SSA-LSTM model are closest to the actual values, and its RMSE, MAE, and MAPE values are all smaller than those of the LSTM and VMD-LSTM models, indicating that the prediction effect of the VMD-SSA-LSTM model is better than that of the LSTM and VMD-LSTM models.
随着经济的持续增长和金融科技的不断发展,个人信贷作为一种满足消费需求的金融工具,其市场规模自然随之扩大。受到经济下行压力、不良贷款行为增加与各种突发变故的影响,个人信贷违约率逐渐上升,一个完善且高效的个人信用评估模型其重要性不言而喻。在信用评估过程中,通过一系列的具体指标和因素去判断个人的信用风险,在庞大的市场规模下,需要巨量的资源投入。本文提出了一种基于稀疏优化的逻辑回归模型,其能在保持一定准确度的情况下快速地得出个人风险评估结果。最后通过真实数据,验证所提出稀疏逻辑回归模型的有效性。With the continuous growth of the economy and the development of financial technology, the market scale of personal credit, as a financial tool to satisfy consumer demand, has naturally expanded. Influenced by the economic downward pressure, the increase of non-performing loan behaviors and various unexpected changes, the default rate of personal credit is gradually rising, and the importance of a perfect and efficient personal credit assessment model is self-evident. In the process of credit assessment, a series of specific indicators and factors are used to judge the credit risk of an individual, which requires a huge amount of resources under a huge market scale. In this paper, a logistic regression model based on sparse optimization is proposed, which can quickly produce individual risk assessment results while maintaining a certain degree of accuracy. Finally, the effectiveness of the proposed sparse logistic regression model is verified by real data.
MEMS传感器在电子商务系统中扮演着至关重要的角色,物联网技术通过连接物理世界与数字世界,为电子商务带来了巨大的创新机遇。它们不仅满足了物联网对于传感器微型化、智能化、低功耗和成本效益的需求,而且随着技术的进步,其应用领域还在不断拓宽。因此MEMS设备的可靠性评估对电子商务技术的发展有着至关重要的作用。MEMS是具有复杂失效模式的系统,本文考虑到系统在运行过程中,受到冲击作用后,其抵御冲击的能力会下降,因此冲击失效的发生概率会增大,即系统硬失效模式的失效阈值会随着冲击次数的增加而进行相应的改变,因此建立了硬失效阈值时变的竞争失效模型,为电子商务的快速发展提供了可靠的评估。MEMS sensors play a crucial role in IoT systems. They not only meet the needs of IoT for miniaturization, intelligence, low power consumption and cost-effectiveness of sensors, but also widen their application areas with the advancement of technology. Therefore, the reliability assessment of MEMS devices plays a crucial role in the development of IoT technology. MEMS is a system with complex failure modes. In this paper, considering that the system’s ability to withstand shocks will be reduced when it is subjected to shocks during operation, the probability of shock failure will increase, i.e., the failure threshold of the system’s hard failure modes will be changed accordingly with the increase in the number of shocks, so a hard failure threshold is established. Therefore, a competitive failure model with time-varying hard failure threshold is established.
创业不仅可以活跃经济、提供就业岗位改善民生,还可以促进协同创新赋能高质量发展。由此,为探究数字化转型背景下创业活跃度的影响因素,本文选择了2011~2021年的全国31个省份的面板数据,从全局的角度出发,综合考虑已有研究中所涉及的22个因素对创业活跃度的影响。首先,考虑到创业活跃度影响因素间的相关性,本文通过利用因子分析对变量进行降维;其次,利用固定效应模型探究各因子对创业活跃度的影响效应;最后,考虑到影响因素众多,为确切地分析各因素对创业活跃度的影响程度,本文利用随机森林对变量重要性进行分析。分析结果显示:数字化水平、社会经济和创业成本显著影响于创业活跃度,影响度最大的绿色金融的发展目前对地方创业活跃起着抑制作用。虽然绿色转型已成为目前企业发展的主要方向,但是由于目前绿色金融市场的目标客户仍是“大型”和“重污”企业,加上较为单一的产品使得小微企业很难在其中找到自己的位置,从而抑制了创业活跃度。本文的结论对于促进数字化转型背景下地方创业活跃度具有重要的启示意义。Entrepreneurship can not only activate the economy, provide jobs and improve people’s livelihood, but also promote collaborative innovation and empower high-quality development. To explore the factors influencing entrepreneurial activity in the context of digital transformation, this study utilizes panel data from 31 provinces across China from 2011 to 2021. It takes a holistic approach, considering the impact of 22 factors from existing research on entrepreneurial activity. Firstly, considering the correlation between factors affecting entrepreneurial activity, this article uses factor analysis to reduce the dimensionality of variables. Secondly, using the fixed-effect model, we explore the impact of various factors on entrepreneurial activity. Finally, considering the numerous influencing factors, in order to accurately analyze the impact of each factor on entrepreneurial activity, this article uses random forest to analyze the importance of variables. The analysis results show that the level of digitalization, socio-economic factors, and entrepreneurial costs significantly affect entrepreneurial activity. The development of green finance, which has the greatest impact, currently has an inhibitory effect on local entrepreneurial activity. Although green transformation has become the main direction of enterprise development, due to the fact that the target customers of the current green financial market are still “large” and “heavy pollution” enterprises, coupled with relatively single products, it is difficult for small and micro enter
暂无评论