基于离散选择模型的推荐系统改进算法

准确地预估用户的点击率,并根据该概率对商品排序以供用户选择在推荐系统领域有着重要的意义。推荐系统中常用的因子分解机等机器学习模型一般只考虑用户选择单个商品的概率,忽略了候选商品之间的相互影响,离散选择模型则考虑将商品候选集作为整体进行考虑。提出了使用深度学习模型来改进离散选择模型,模型使用相对特征层、注意力机制等网络结构帮助深度学习模型进行不同商品间的特征比较,研究结果表明引入离散选择模型的深度学习模型表现优于梯度提升决策树、因子分解机等模型。     

Machine Learning Classification Model Comparison

Machine learning models are boosting Artificial Intelligence applications in many domains, such as automotive, finance and health care. This is mainly due to their advantage, in terms of predictive accuracy, with respect to classic statistical models. However, machine learning models are much less explainable: less transparent, less interpretable. This paper proposes to improve machine learning models, by proposing a model selection methodology, based on Lorenz Zonoids, which allows to compare them in terms of predictive accuracy significant gains, leading to a selected model which maintains accuracy while improving explainability. We illustrate our proposal by means of simulated datasets and of a real credit scoring problem. The analysis of the former shows that the proposal improves alternative methods, based on the AUROC. The analysis of the latter shows that the proposal leads to models made up of two/three relevant variables that measure the profitability and the financial leverage of the companies asking for credit.     

Horseshoe Regularisation for Machine Learning in Complex and Deep Models1

Since the advent of the horseshoe priors for regularisation, global–local shrinkage methods have proved to be a fertile ground for the development of Bayesian methodology in machine learning, specifically for high-dimensional regression and classification problems. They have achieved remarkable success in computation and enjoy strong theoretical support. Most of the existing literature has focused on the linear Gaussian case; for which systematic surveys are available. The purpose of the current article is to demonstrate that the horseshoe regularisation is useful far more broadly, by reviewing both methodological and computational developments in complex models that are more relevant to machine learning applications. Specifically, we focus on methodological challenges in horseshoe regularisation in non-linear and non-Gaussian models, multivariate models and deep neural networks. We also outline the recent computational developments in horseshoe shrinkage for complex models along with a list of available software implementations that allows one to venture out beyond the comfort zone of the canonical linear regression problems.     

A decomposition clustering ensemble learning approach for forecasting foreign exchange rates

A decomposition clustering ensemble (DCE) learning approach is proposed for forecasting foreign exchange rates by integrating the variational mode decomposition (VMD), the self-organizing map (SOM) network, and the kernel extreme learning machine (KELM). First, the exchange rate time series is decomposed into N subcomponents by the VMD method. Second, each subcomponent series is modeled by the KELM. Third, the SOM neural network is introduced to cluster the subcomponent forecasting results of the in-sample dataset to obtain cluster centers. Finally, each cluster's ensemble weight is estimated by another KELM, and the final forecasting results are obtained by the corresponding clusters' ensemble weights. The empirical results illustrate that our proposed DCE learning approach can significantly improve forecasting performance, and statistically outperform some other benchmark models in directional and level forecasting accuracy.     

Forecasting third-party mobile payments with implications for customer flow prediction

Forecasting customer flow is key for retailers in making daily operational decisions, but small retailers often lack the resources to obtain such forecasts. Rather than forecasting stores’ total customer flows, this research utilizes emerging third-party mobile payment data to provide participating stores with a value-added service by forecasting their share of daily customer flows. These customer transactions using mobile payments can then be utilized further to derive retailers’ total customer flows indirectly, thereby overcoming the constraints that small retailers face. We propose a third-party mobile-payment-platform centered daily mobile payments forecasting solution based on an extension of the newly-developed Gradient Boosting Regression Tree (GBRT) method which can generate multi-step forecasts for many stores concurrently. Using empirical forecasting experiments with thousands of time series, we show that GBRT, together with a strategy for multi-period-ahead forecasting, provides more accurate forecasts than established benchmarks. Pooling data from the platform across stores leads to benefits relative to analyzing the data individually, thus demonstrating the value of this machine learning application.     

Can media and text analytics provide insights into labour market conditions in China?

The official Chinese labour market indicators have been seen as problematic given their small cyclical movement and their only partial capture of the labour force. In our paper, we build a monthly Chinese labour market conditions index (LMCI) using text analytics applied to Mainland Chinese-language newspapers over the period from 2003 to 2017. We use a supervised machine learning approach by training a support vector machine classification model. The information content and the forecast ability of our LMCI are tested against official labour market activity measures in wage and credit growth estimations. Surprisingly, one of our findings is that the much-maligned official labour market indicators do contain information. However, their information content is not robust and, in many cases, our LMCI can provide forecasts that are significantly superior. Moreover, regional disaggregation of the LMCI illustrates that labour conditions in the export-oriented coastal region are sensitive to export growth, while those in inland regions are not. This suggests that text analytics can, indeed, be used to extract useful labour market information from Chinese newspaper articles.     

Machine learning advances for time series forecasting

In this paper, we survey the most recent advances in supervised machine learning (ML) and high-dimensional models for time-series forecasting. We consider both linear and nonlinear alternatives. Among the linear methods, we pay special attention to penalized regressions and ensemble of models. The nonlinear methods considered in the paper include shallow and deep neural networks, in their feedforward and recurrent versions, and tree-based methods, such as random forests and boosted trees. We also consider ensemble and hybrid models by combining ingredients from different alternatives. Tests for superior predictive ability are briefly reviewed. Finally, we discuss application of ML in economics and finance and provide an illustration with high-frequency financial data.     

Statistical learning and exchange rate forecasting

This study uses innovative tools recently proposed in the statistical learning literature to assess the capability of standard exchange rate models to predict the exchange rate in the short and long runs. Our results show that statistical learning methods deliver remarkably good performance, outperforming the random walk in forecasting the exchange rate at different forecasting horizons, with the exception of the very short term (a period of one to two months). These results were robust across countries, time, and models. We then used these tools to compare the predictive capabilities of different exchange rate models and model specifications, and found that sticky price versions of the monetary model with an error correction specification delivered the best performance. We also explain the operation of the statistical learning models by developing measures of variable importance and analyzing the kind of relationship that links each variable with the outcome. This gives us a better understanding of the relationship between the exchange rate and economic fundamentals, which appears complex and characterized by strong non-linearities.     

基于省域视角的信用风险模型适应性分析

巴塞尔新资本协议在鼓励银行采用内部评级法评估信用风险以提取资本准备的同时也强化了各国监管机构对内部评级模型绩效检验与审查的要求.CreditMetrics和CreditRisk+是银行业信用风险评估的基准模型.从建模的数学方法看,CreditRisk+是基于违约的判断,而CreditMetrics则是根据等级变化评价.利用江苏省银监局的相关统计数据对信用风险评估模型进行参数特性审查与绩效检验,结果显示这两类常用模型都可以在江苏的商业银行经营实践中稳定地实现根据信贷组合的实际风险状况进行内部资本配置这一目标.     

Predictive analysis and modelling football results using machine learning approach for English Premier League

The introduction of artificial intelligence has given us the ability to build predictive systems with unprecedented accuracy. Machine learning is being used in virtually all areas in one way or another, due to its extreme effectiveness. One such area where predictive systems have gained a lot of popularity is the prediction of football match results. This paper demonstrates our work on the building of a generalized predictive model for predicting the results of the English Premier League. Using feature engineering and exploratory data analysis, we create a feature set for determining the most important factors for predicting the results of a football match, and consequently create a highly accurate predictive system using machine learning. We demonstrate the strong dependence of our models’ performances on important features. Our best model using gradient boosting achieved a performance of 0.2156 on the ranked probability score (RPS) metric for game weeks 6 to 38 for the English Premier League aggregated over two seasons (2014–2015 and 2015–2016), whereas the betting organizations that we consider (Bet365 and Pinnacle Sports) obtained an RPS value of 0.2012 for the same period. Since a lower RPS value represents a higher predictive accuracy, our model was not able to outperform the bookmaker’s predictions, despite obtaining promising results.     

存在规模效应的再制造回收渠道决策模型

在第三方回收存在规模效应的情况下,文中建立了再制造闭环供应链的回收渠道决策模型。比较了三种回收渠道中回收率、零售价、批发价以及制造商和零售商的利润关系,以及通过算例分析了规模效应对第三方回收率以及供应链上各成员利润的影响。结果表明：规模效应较大时,第三方回收渠道优于制造商和零售商回收渠道;在第三方回收模型中,回收率随着规模效应的减小而减小;规模效应较大时,第三方分得最多的再制造利润,其次是制造商,但随着规模效应的减小,利润逐渐转移到零售商那里。     

基于Stacking模型融合的用户购买行为预测研究

在大数据时代背景下,如何利用大量的销售数据精准预测顾客未来需求,成为企业制定客户管理和库存管理决策的一个重要问题。目前关于用户购买行为预测的研究中很少能够预测用户具体的购买时间。基于已有的销售数据,提出了基于机器学习和Stacking集成的综合预测模型预测用户的购买行为,即未来是否购买及其购买时间。将模型应用在一家大型连锁零售企业的需求预测中,并对方法的有效性进行评估。结果表明,基于Stacking集成的融合模型对预测用户未来是否购买具有最佳性能,准确率达85%,AUC值达到0.928;LightGBM集成算法在预测用户购买时间时具有最优性能,相比于融合模型提升了5.5%的预测性能;融合模型+LightGBM算法的组合相比于均使用融合模型提升了9.4%的预测性能。     

Homogeneous semi-Markov reliability models for credit risk management*

Abstract The credit risk problem is one of the most important issues of modern financial mathematics. Fundamentally it consists in computing the default probability of a company going into debt. The problem can be studied by means of Markov transition models. The generalization of the transition models by means of homogeneous semi-Markov models is presented in this paper. The idea is to consider the credit risk problem as a reliability problem. In a semi-Markov environment it is possible to consider transition probabilities that change as a function of waiting time inside a state. The paper also shows how to apply semi-Markov reliability models in a credit risk environment. In the last section an example of the model is provided. Mathematics Subject Classification (2000): 60K15, 60K20, 90B25, 91B28 Journal of Economic Literature Classification: G21, G33     

银行不良贷款违约损失率结构特征研究

本文对中国银行业面临的信用风险违约损失率(LGD)展开研究,以温州某商业银行不良贷款数据为样本,通过描述性统计,对LGD的结构特征:信用风险暴露规模特征、期限特征、地域特征以及担保特征等进行了详细分析。结果表明LGD与风险暴露规模呈负相关,LGD与贷款期限呈正相关,不同地域、不同担保方式的违约贷款其LGD差异性显著。以上这些结论可为商业银行信用风险管理、信贷投放导向以及信用风险监管提供现实帮助。     

News will tell: Forecasting foreign exchange rates based on news story events in the economy calendar

The paper proposes a novel approach to predict intraday directional-movements of currency-pairs in the foreign exchange market based on news story events in the economy calendar. Prior work on using textual data for forecasting foreign exchange market developments does not consider economy calendar events. We consider a rich set of text analytics methods to extract information from news story events and propose a novel sentiment dictionary for the foreign exchange market. The paper shows how news events and corresponding news stories provide valuable information to increase forecast accuracy and inform trading decisions. More specifically, using textual data together with technical indicators as inputs to different machine learning models reveals that the accuracy of market predictions shortly after the release of news is substantially higher than in other periods, which suggests the feasibility of news-based trading. Furthermore, empirical results identify a combination of a gradient boosting algorithm, our new sentiment dictionary, and text-features based-on term frequency weighting to offer the most accurate forecasts. These findings are valuable for traders, risk managers and other consumers of foreign exchange market forecasts and offer guidance how to design accurate prediction systems.     

Credit ratings of Chinese online loan platforms based on factor scores and K-means clustering algorithm

The rapid development of Chinese online loan platforms (OLPs), as well as their risks, has attracted widespread attention, increasing the demand for a complete credit rating mechanism. The present study establishes a credit rating indicator system for 130 mainstream Chinese OLPs that combines 12 quantitative metrics of online loan operations similar to commercial bank credit rating indicators, including platform transaction volume and average expected rate of return. We also consider two qualitative indicators of online loan background, namely platform background and guarantee mode, that reflect Chinese characteristics. Subsequently, a factor analysis was conducted to reduce the 14 indicators’ dimensions. The loads of the rating indicators in the resulting rotating component matrix were refined into an OLP operation scale factor, fund dispersion factor, security factor, and profitability factor. Finally, a K-means clustering algorithm was employed to cluster the factor scores of each OLP, thereby obtaining credit rating results. The empirical results indicate that the proposed machine learning–based credit rating method effectively provides early warnings of problem platforms, yielding more accurate credit ratings than those provided by two mainstream online loan rating websites in China, namely, Wangdaitianyan and Wangdaizhijia.     

A novel two-stage method for well-diversified portfolio construction based on stock return prediction using machine learning

The performance of portfolio model can be improved by introducing stock prediction based on machine learning methods. However, the prediction error is inevitable, which may bring losses to investors. To limit the losses, a common strategy is diversification, which involves buying low-correlation stocks and spreading the funds across different assets. In this paper, a diversified portfolio selection method based on stock prediction is proposed, which includes two stages. To be specific, the purpose of the first stage is to select diversified stocks with high predicted returns, where the returns are predicted by machine learning methods, i.e. random forest (RF), support vector regression (SVR), long short-term memory networks (LSTM), extreme learning machine (ELM) and back propagation neural network (BPNN), and the diversification level is measured by Pearson correlation coefficient. In the second stage, the predictive results are incorporated into a modified mean–variance (MMV) model to determine the proportion of each asset. Using China Securities 100 Index component stocks as study sample, the empirical results demonstrate that the RF+MMV model achieves better results than similar counterparts and market index in terms of return and return–risk metrics.     

Bagging方法在基于日常交易数据解决中小企业的信用评价问题中的应用

关于信用风险评价问题至今已经做了很多研究,各种信用评价模型与方法也已被开发。但是这些模型与方法几乎都是基于财务数据、股票价格或风险调研机构发表的各种调查结果。因为几乎所有的中小企业的财务数据都是非公开的,至今开发的信用评价模型与方法都不免成为无米之炊。为此,本文提出了一个新的途径,只需要根据销售额、顾客付款额、拖欠款额等日常业务处理数据来评价顾客企业的信用度。本文提出一个应用Sagging方法评价顾客信用的系统,其目的在于解决由于异常顾客数比正常顾客要少很多而带来的问题,提高分辨异常顾客的能力。本文所提出的信用评价系统将应用到一个实际企业的信用评价问题中,借此来验证系统的性能和效果。     

Evidence on the effect of credit counseling on mortgage loan default by low-income households

Research on the effectiveness of credit counseling is surprisingly scarce given its widespread use and given that it has been around for at least three decades. This paper studies the effects of counseling on default by adopting an option-based approach to mortgage termination. Data come from a counseling program developed as result of the collaborative efforts of a large Midwest bank, Community Churches, and a local community development company implemented during the 1992–1996 period. We find some evidence that counseled borrowers defaulted less often than non-counseled borrowers and that counseling affects optimal exercise of the default option.