Information and Knowledge as Antecedents of Consumer Attitudes and Intentions to Buy and Recommend Fair-Trade Products

The authors propose and empirically test a causal model to understand how the availability of fair-trade information and consumer knowledge about this issue affect consumers’ attitudes and intentions toward fair-trade products. The model is built upon the attitude-behavior paradigm and the premises of agency theory. It is tested through structural equation modeling with a sample of 292 Spanish consumers. The findings are that consumers do not have good knowledge about fair trade and that this is significantly determined by the lack of information about this in the market. It is also observed that consumers’ perceptions about the availability of fair-trade information have negative effects on their concern about this issue and that such information as is available is not effective in reducing consumer skepticism. The research represents an extension of previous fair-trade literature because the role of information and communication in improving consumer attitudes and buying intentions has rarely been explored in the case of ethical products.     

Testing the asymmetric effects of exchange rate pass‐through in BRICS countries: Does the state of the economy matter?

This paper investigates not only the question of whether there is exchange rate pass‐through (ERPT) but also the extent to which the pass‐through is asymmetric or state‐dependent in the BRICS countries. Using monthly data from 1999:M1 to 2019:M12 and non‐linear smooth transition vector autoregressive (STVAR) model, our results provide evidence of period‐specific ERPT between the upper and lower regime periods, governed by the selected transition variables. The results further suggest that the pass‐through of exchange rate is higher when the economy is experiencing large appreciations and expansions as well as large depreciations and recessions. Theimplication for these findings is that ERPT is strongly affected by the state of the economy.     

基于改进Transformer模型的文本摘要生成方法

传统的文本摘要方法，如基于循环神经网络和Encoder-Decoder框架构建的摘要生成模型等，在生成文本摘要时存在并行能力不足或长期依赖的性能缺陷，以及文本摘要生成的准确率和流畅度的问题。对此，提出了一种动态词嵌入摘要生成方法。该方法基于改进的Transformer模型，在文本预处理阶段引入先验知识，将ELMo(Embeddings from Language Models)动态词向量作为训练文本的词表征，结合此词对应当句的文本句向量拼接生成输入文本矩阵，将文本矩阵输入到Encoder生成固定长度的文本向量表达，然后通过Decoder将此向量表达解码生成目标文本摘要。实验采用Rouge值作为摘要的评测指标，与其他方法进行的对比实验结果表明，所提方法所生成的文本摘要的准确率和流畅度更高。     

基于机器学习的斜坡堤越浪量预测方法研究

针对斜坡堤越浪量预测方法,分别建立集成神经网络(ensemble neural network,ENN)、随机森林(random for-eset,RF)和支持向量回归机(suppport vector regression,SVR)3种机器学习模型对斜坡堤越浪量进行预测,并利用决定系数R2和均方根误差RMSE来评估模型性能.最后,对3种模型的性能进行分析.结果显示,集成神经网络模型的决定系数R2和均方根误差RM S E分别约为0.96和0.0018,随机森林模型的决定系数R2和均方根误差RMSE分别约为0.97和0.0014,支持向量回归机模型的决定系数R2和均方根误差RMSE分别约为0.94和0.002.对比发现,3种模型的决定系数都达到0.9以上,都具有较高的预测精度,随机森林相比其他两个模型精度更高.     

Machine learning for financial transaction classification across companies using character-level word embeddings of text fields

An important initial step in accounting is mapping financial transfers to the corresponding accounts. We devised machine-learning-based systems that automate this process. They use word embeddings with character-level features to process transaction texts. When considering 473 companies independently, our approach achieved an average top-1 accuracy of 80.50%, outperforming baselines that exclude the transaction texts or rely on a lexical bag-of-words text representation. We extended the approach to generalizes across companies and even across different corporate sectors. After standardization of the account structures and careful feature engineering, a single classifier trained on 44 companies from 28 sectors achieved a test accuracy of more than 80%. When trained on 43 companies and tested on the remaining one, the system achieved an average performance of 64.62%. This rate increased to nearly 70% when considering only the largest sector.     

Perceived hygiene attributes in the hotel industry: customer retention amid the COVID-19 crisis

The COVID-19 pandemic has caused a crisis in the hotel industry worldwide, but few studies have suggested methods to retain customers. This study proposes hygiene management as a means to minimize the indirect damage from COVID-19 to the hotel industry. It identifies perceived hygiene attributes and explores their influence on hotel image, word of mouth, and revisit intentions. This study identifies and validates three types of perceived hygiene attributes through qualitative and quantitative methods. It uses structural equation modeling to validate hypotheses and concludes that there are significant relationships of influence between the proposed variables. This study provides important and meaningful insights into hotel image and customer behavior through perceived hygiene attributes.     

Combining wavelet decomposition with machine learning to forecast gold returns

This paper combines the discrete wavelet transform with support vector regression for forecasting gold-price dynamics. The advantages of this approach are investigated using a relatively small set of economic and financial predictors. I measure model performance by differentiating between a statistically-motivated out-of-sample forecasting exercise and an economically-motivated trading strategy. Disentangling the predictors with respect to their time and frequency domains leads to improved forecasting performance. The results are robust compared to alternative forecasting approaches. My findings on the relative importances of such wavelet decompositions suggest that the influences of short-term and long-term trends are not stable over the full evaluation period.     

A multivariate distance nonlinear causality test based on partial distance correlation: a machine learning application to energy futures

This paper proposes a multivariate distance nonlinear causality test (MDNC) using the partial distance correlation in a time series framework. Partial distance correlation as an extension of the Brownian distance correlation calculates the distance correlation between random vectors X and Y controlling for a random vector Z. Our test can detect nonlinear lagged relationships between time series, and when integrated with machine learning methods it can improve the forecasting power. We apply our method as a feature selection procedure and combine it with the support vector machine and random forests algorithms to study the forecast of the main energy financial time series (oil, coal, and natural gas futures). It shows substantial improvement in forecasting the fuel energy time series in comparison to the classical Granger causality method in time series.     

一种基于Radon变换的微表情识别算法

微表情是人们处在一些与平时生活环境不同的高强度环境下试图控制和掩饰的情感表现，也是一种不曾意识到的瞬时脸部表情,持续时间短，强度弱。为了提高其准确率，提出了基于Radon变换的微表情识别算法。首先，对数据库中的视频序列进行灰度归一化、尺寸归一化和二维主成分分析法（Two-dimensional Principal Component Analysis，2DPCA）降维预处理，使用光流法对降维后图像提取运动特征；然后使用Radon变换算法对光流图像进行处理，得到对应微表情的特征值和特征图像；最后使用支持向量机进行微表情分类识别。实验结果表明，使用Radon变换后得到的微表情特征图像得到了较好的识别效果，在微表情数据集CASME和CASMEⅡ上识别率分别为81.48%和82.17%，通过与选取的其他方法对比说明了该方法具有更好的识别性能。     

结合新型模糊支持向量机和证据理论的多传感器水质数据融合

在多传感器水质数据融合领域，证据理论是有效的数据融合方法之一，但基本概率分配一般不易确定，从而使数据融合能力难以有效发挥。支持向量机是统计学习理论之上的高级分类算法，具有普适性和全局优化等特点，但输出的基本概率分配有待进一步提高。提出了一种基于证据理论和新型模糊支持向量机相结合的数据融合方法，通过建立基于分类超平面距离的模糊隶属度，训练模糊支持向量机提高传统支持向量机的基本概率分配，并结合证据理论进行海河水质数据融合。通过证据理论分别结合支持向量机和模糊综合评价法与上述方法进行对比实验，经精度、平均绝对百分误差、均方根误差等指标验证，精度提高10.5%，表明所提方法是一种可靠的多传感器的水质融合方法，较其他方法具有更高的融合精度。