基于神经网络的综合方法的研究

对神经网络与其它智能控制技术相结合进行了综述,主要有神经网络与模糊逻辑技术、小波分析、可拓理论、遗传算法和混沌等技术相互结合从而形成新型神经网络。进而概述了小波网络、模糊网络、模糊小波神经网络、遗传神经网络、混沌神经网络和开拓神经网络等模型的研究现状,并展望了今后的发展趋势。     

The true story of wine and cloth, or: building blocks of an evolutionary political economy of international trade

Standard trade theory is a theory about the structure of international production which emerges from an international arbitrage equilibrium. It is not a theory about the activity of trading which is simply taken for granted or which is treated implicitly in the concept of exogenous trade costs. This paper proposes an alternative evolutionary framework based on networks as structures of non-price interactions into which price/quantity-interactions are embedded. These networks manifest different levels with specific problems of coordination and communication, i.e. levels of market transactions, of transaction-enabling transactions, of markets for market access rights, and of the respective transaction-enabling transactions. Furthermore, the theory is based on the analysis of capabilities to trade, resting upon competitive advantage, which cannot be imitated. The exploitation of competitive advantage presupposes the capability to control network interactions, identified as social capital. Finally, the security of market access reflects power balances among countries.

基于IPSO-BP神经网络的铁路客运量预测

在分析有关铁路客运量预测方法的基础上,针对BP神经网络模型存在的不足,提出基于粒子群优化算法(PSO)优化BP神经网络的参数,即改进的PSO方法(IPSO)。以我国1990—2007年的铁路客运量为研究对象,确定输入样本和输出样本,以及训练集和测试集,建立基于IPSO的BP神经网络优化模型预测铁路客运量。预测结果表明,IPSO-BP网络的算法训练时间短,收敛速度快,预测精度高。     

User-oriented design for the optimal combination on product design

This paper presents a new approach of user-oriented design for transforming users’ perception into product elements design. An experimental study on mobile phones is conducted to examine how product form and product color affect product image individually and as a whole. The concept of Kansei Engineering is used to extract the experimental samples as a data base for Quantitative Theory Type I and neural networks (NNs). The result of numerical analysis suggests that mobile phone makers need to provide various product colors to attract users, in addition to product forms. This paper demonstrates the advantage of using NNs for determining the optimal combination of product form and product color, particularly if the product into design elements. Based on the analysis of NNs, we can use 72 representative product colors of each mobile phone to develop a product color data base consisting of 16777216 (=256×256×256, True-Color model) colors with the associated product image. The design data base provides useful insights to save any amount of money and time for the new product development. The product designers can input a product image to work out an adequate color on a mobile phone. Furthermore, the design data base can be used, in conjunction with computer-aided design system or virtual reality technology, to build a 3D model for facilitating the design process of mobile phones. Although, the mobile phones are chosen as the object of the experimental study, this approach can be applied to other products with various design elements.     

基于模糊神经网络的商品价格预测模型

为了方便地预测大型商品的价格,建立了一个模糊神经网络预测模型。该模型能够通过分析影响商品价格的各种参数及历史数据来评定一个新商品的大致价格水平。同时,在比较了FNN与ANN的基础上得出了FNN的优势。     

基于EEMD-BP方法的城市轨道交通进站客流短期预测

城市轨道交通站点客流量是评价其服务水平和实现城市轨道交通资源有效配置的基础数据和依据。针对城市轨道交通站点进站客流量序列波动复杂的问题,构建基于EEMD-BP方法的城市轨道交通进站客流短期预测模型,通过对城市轨道交通站点的日间分时进站客流序列进行模态分解,并对分解的分量进行筛选和识别,探究进站客流的日间波动影响因素,实现对短期客流的合理预测。以广州珠江新城站短期客流预测为例,验证该组合模型在提高客流预测方面具有有效性,为城市轨道交通线网规划和运营管理提供客流预测依据。     

Credit risk interconnectedness: What does the market really know?

We analyze the relation between market-based credit risk interconnectedness among banks during the crisis and the associated balance sheet linkages via funding and securities holdings. For identification, we use a proprietary dataset that has the funding positions of banks at the bank-to-bank level for 2006–2013 in conjunction with investments of banks at the security level and the credit register from Germany. We find asymmetries both cross-sectionally and over time: when banks face difficulties to raise funding, the interbank lending affects market-based bank interconnectedness. Moreover, banks with investments in securities related to troubled classes have a higher credit risk interconnectedness. Overall, our results suggest that market-based measures of interdependence can serve well as risk monitoring tools in the absence of disaggregated high-frequency bank fundamental data.     

Developing an organizational typology of criminals in the meat supply chain

The European Horsemeat Scandal of 2013 highlighted the increasing organization and sophistication of the contemporary food criminal. This study aims to develop a more sophisticated understanding of the typology of the food criminal in terms of their modus operandi and how individuals and organized crime groups develop criminal business models and networks in the context of meat supply. This research initiates a synthesized literature review across the seemingly disparate academic disciplines of food and agricultural policy, business theory and criminology in order to characterize the modes of operation at work in such networks. A conceptual framework is developed that considers the actors and drivers involved in criminal activity using the meat supply chain as an example.     

Non-linearities in financial bubbles: Theory and Bayesian evidence from S&P500

The modeling process of bubbles, using advanced mathematical and econometric techniques, is a young field of research. In this context, significant model misspecification could result from ignoring potential non-linearities. More precisely, the present paper attempts to detect and date non-linear bubble episodes. To do so, we use Neural Networks to capture the neglected non-linearities. Also, we provide a recursive dating procedure for bubble episodes. When using data on stock price-dividend ratio S&P500 (1871.1–2014.6), employing Bayesian techniques, the proposed approach identifies more episodes than other bubble tests in the literature, while the common episodes are, in general, found to have a longer duration, which is evidence of an early warning mechanism (EWM) that could have important policy implications.     

Predicting macro-financial instability – How relevant is sentiment? Evidence from long short-term memory networks

This paper examines the relevance of sentiment in predicting overall financial system instability using long-run short-term memory networks. Weekly data on the US financial system, consumer sentiment, producer sentiment, and investor sentiment is collected from 21 January 1994 to 27 December 2019, and different models are developed to predict the one-week-ahead levels of financial stress in the US financial system. We find that models using sentiment indices outperform those relying solely on historical financial stress and risk data. This result is robust to comparisons with an alternative deep learning method and out-of-sample predictions. It constitutes an argument in favor of behavioral finance and Minsky’s (Knell, 2015) financial instability hypothesis against the Efficient Market Hypothesis. As it concretely identifies the main indicators for predicting US financial stress one week in advance, the study provides relevant recommendations for policymakers and investors in terms of macroprudential policies and portfolio management.