基于集成学习的多式联运货运量预测模型构建

货运量精准预测是多式联运网络高效协同发展的重要基础,货运量时变性强、数据多样性缺失是实现精准货运量预测的问题所在。基于此,通过挖掘货物运输量(集装箱)的时间变化特征,构建初始相关时间特征输入集,结合斯皮尔曼相关性系数分布,采用Bagging+BP集成学习方法训练多个弱分类器,最终组合获取高精度的强学习模型。以南京龙潭港为例,对自回归移动平均模型(ARIMA)、Bagging+BP集成学习网络以及长短时记忆神经网络(LSTM)三种模型进行评价,实验结果表明,相比于其他模型,提出的Bagging+BP集成学习网络预测性能良好,有一定的实用价值。     

基于Stacking策略和报童模型的补货决策方法

针对市场中供需关系的不确定性会影响商家的补货决策问题,本文提出了基于Stacking融合预测及报童模型的智能补货决策方法,并将模型应用于某商店。结果显示融合模型的MSE,RMSE,MAE显著低于单个基模型,说明预测模型的精度得到了改进。本文所设计的方法可辅助商家制定营销策略,减少产品滞销或供不应求带来的损失。     

基于用户分类的余额宝资金流预测研究

以余额宝为代表的互联网金融作为一种创新的金融服务模式对资金流的预测提出了新的要求。本文以余额宝资金流数据入手,把用户分类融入时间序列算法,提出了基于用户分类的资金流预测模型。通过实证分析,该模型与传统的时间序列模型相比提高了预测的准确性。     

基于PSO优化LSSVM在基坑开挖过程中周围建筑物沉降预测研究

本文提出一种用粒子群优化算法来确定LSSVM参数的方法。该方法是在对LSSVM进行分析的基础上,融合PSO的群搜索特征来提高LSSVM预测精度。文章最后采用昆明市某基坑周围建筑物沉降数据对此模型进行了验证,并与其他算法进行了对比分析,计算结果表明用该模型进行沉降预测相比其他算法具有较快的收敛速度和更高的预测精度。     

基于神经网络集成方法的公司盈利预测研究

盈余预测具有引导投资者投资行为的作用,因此受到投资者的广泛重视。然而,国内对公司未来盈利进行预测的研究还相当少。提出了以决策树作为基分类器,采用集成学习方法,利用上市某公司2001至2005年的财务数据对该上市公司在2006年的盈利状况进行预测研究。首先,采用有放回的随机抽样技术分别从训练样本和测试样本中产生50个训练子集和1个测试集;然后利用决策树,采用CHAID算法对50个训练子集分别进行训练,得到50个基决策树分类器;通过采用Bagging方法,构建决策树集成模型。所得到的集成模型在测试集上的分类准确率达到96％以上,通过比较由不同数目的基分类器构成的集成模型和单个分类器的预测准确率,证明了该集成模型的预测准确率高且稳定。     

遗传算法优化的BP神经网络对汽车销售的预测

为了使BP神经网络拟合复杂函数的能力得到进一步提升,本文通过改变BP神经网络结构的权值和阈值,从而进一步优化遗传算法,并将此设计的模型应用到预测汽车销售系统之中。为了比较它与传统BP神经算法预测效果,我们进行了预测仿真,从中可以比较两者准确性高低与否。仿真结果表明改进后的算法对存在较明显线性相关性的数据时有更好的拟合能力和更高的预测精度。     

整车物流中销售模型的统计及预测

根据整车生产销售的复杂性,以汽车整车的年销售量为依据,建立整车销售的Logistic模型,并说明了利用该模型研究整车销售情况的优越性。文中依据以往的数据对我国未来几年的整车销售量作了短期预测,为汽车生产商对整车的生产行为作出及时地调整提供了有利数据,继而为第三方物流公司也提供了可靠的科学依据,从而有助于实现整车物流的最优化。     

电子商务平台顾客感知价值对购买行为及企业未来销售的影响

在全球经济一体化深入发展的今天,电子商务凭借其不受地域和时间限制、运行成本低、效率高、交易快捷方便等特点,越来越受到企业和个人的青睐。电子商务企业要想在日渐激烈的竞争中获得更多的利益,必须分析顾客的价值需求,通过及时响应顾客需求,影响顾客购买行为,维系与顾客的关系,进而培养顾客忠诚。本文旨在电子商务的环境下,研究顾客感知价值对购买意向及企业未来销售的影响。首先,通过文献研究提出研究假设;其次,利用带有"时间窗"的固定样本数据采集技术以及三阶段最小二乘法对调查数据进行收集和分析;再次,进行假设检验;最后,研究结果发现感知价值的几个维度即社会价值、功能价值、情感价值和程序价值均对购买意向产生正向的影响,并且,购买意向对企业未来的销售有积极的正向的影响,同时过去的销售额对未来的销售额有很好的预测作用。     

基于注意力机制的LSTM股价趋势预测研究

针对中国股票市场,提出了一种基于注意力机制的LSTM股价趋势预测模型。选取42只中国上证50从2009年到2017年的股票数据为实验对象,根据股票市场普遍认可的经验规则,分别对每个技术指标进行量化处理得到股票涨跌的趋势数据,并和交易数据混合作为预测模型的输入,然后使用基于注意力机制的LSTM模型提取股价趋势特征进行预测。实验结果表明:引入股票离散型趋势数据到预测模型中,能够在已有交易数据和技术指标的基础上提升预测精确度,与传统的机器学习模型SVM和单一的LSTM模型相比,基于注意力机制的LSTM模型具有更好的预测能力。     

企业提升预测大宗商品价格精准度的方法研究--基于时频递归与非递归分解的混合价格预测模型的运用

大宗商品交易在我国经济循环中占比极高,但国内众多企业由于没有选取世界大宗商品价格预测最优方法,造成巨额损失屡见不鲜。在时间序列信号的递归于非递归时频分析分析的基础上,构建传统自回归移动平均模型和新兴神经网络的新复合模型对商品期货价格进行预测。使用墨西哥湾沿岸煤油航空燃料价格历史数据作为建模时间序列数据,以此建立航空燃料价格动态预测模型。通过对比,从模型的预测结果来看,结合时频非递归分解的预测模型在预测精度上优于基于递归分解的复合模型优于单一预测模型,可以有利于提升企业对世界大宗商品价格预测的精准度,从而控制成本,有效地降低风险。     

Are green consumers really green? Exploring the factors behind the actual consumption of organic food products

The increasing interest in sustainable consumption has lead several scholars to investigate the determinants that drive the consumption of organic food. Most of this research is based on consumers' self‐reports of their purchasing behavior by exploring declared behavioral intentions. There is a lack of understanding concerning the determinants of organic food consumption based on actual purchasing behavior. To fill this gap, this study is based on a combination of actual purchasing data and self‐reported data from a sample of 79 Italian consumers. The determinants of organic food consumption are explored by analyzing the effects of subjective norms, attitude, perceived behavioral control, intention to buy, organic knowledge, and health consciousness on actual purchasing behavior. Our results suggest that actual purchasing behavior is positively influenced by intention to buy and negatively by subjective norms. Although attitude towards buying organics is positively affected by health consciousness and perceived behavioral control, consumer knowledge about organics is found to influence purchase intentions. Theoretical and managerial implications, along with avenues for future research, are discussed.     

A solution for M5 Forecasting - Uncertainty: Hybrid gradient boosting and autoregressive recurrent neural network for quantile estimation

This paper presents our 13th place solution to the M5 Forecasting - Uncertainty challenge and compares it against GoodsForecast’s second-place solution. This challenge aims to estimate the median and eight other quantiles of various product sales in Walmart. Both solutions handle the predictions of median and other quantiles separately. Our solution hybridizes LightGBM and DeepAR in various ways for median and quantile estimation, based on the aggregation levels of the sales. Similarly, GoodsForecast’s solution also utilized a hybrid approach, i.e., LightGBM for point estimation and a Histogram algorithm for quantile estimation. In this paper, the differences between the two solutions and their results are highlighted. Despite our solution only taking 13th place in the challenge with the competition metric, it achieves the lowest average rank based on the multiple comparisons with the best (MCB) test which implies the most accurate forecasts in the majority of the series. It also indicates better performance at the product-store aggregation level which comprises 30,490 (71.2% of all) series compared to most teams.     

Joint inventory and transshipment decisions with consumer behavioral heterogeneity

Transshipment is an effective method for reducing mismatches between supply and demand among retailers. Consumers attempting to purchase out-of-stock items may wait for transshipment, purchase at another store, or choose not to buy. In this paper, the consumer behavioral heterogeneity is characterized using the transshipment request rate and consumer switching rate. It affects the replenishment and transshipment decisions, as well as the system profits. The inventory replenishment and transshipment decisions are studied in both centralized and decentralized two-location inventory systems with consumer behavioral heterogeneity. We characterize optimal replenishment decisions in a centralized system, prove the existence of a unique Nash equilibrium in a decentralized system with a specific demand distribution, and determine the coordinating transshipment price for some decentralized systems. In numerical studies, the performance of various systems is analyzed for consumers with identical (symmetric) or differing (asymmetric) behavior between retailers. For scenarios with symmetric consumers, a higher transshipment request rate and consumer switching rate resulted in increased total profit in all systems. For scenarios with asymmetric consumers, the retailer with the higher consumer switching rate should reduce ordering in a centralized system but increase ordering in a decentralized system. Moreover, the retailer with the higher transshipment request rate reduces ordering to increase profit, whereas the other retailer increases order quantity yet earns less profit.     

旅游中心城市旅游纪念品购买行为研究——基于北京市的调查数据分析

目前,我国旅游纪念品销售收入远远落后于发达国家。绝大多数旅游者都希望能在旅游地买些纪念品,但是我们经常听到游客抱怨在旅游景区没什么旅游纪念品可买。而景区内商家也抱怨旅游纪念品的生意难做,看的多,买的少。一边是巨大的市场需求得不到满足,一边是供给者抱怨需求不足,这种情况在旅游景区普遍存在。因此,分析旅游纪念品购买行为影响因素,提出我国旅游纪念品市场的改进措施,对于推动旅游纪念品的产业化,促进旅游业的迅速发展具有重要意义。     

What you see is meaningful: Does green advertising change the intentions of consumers to purchase eco-labeled products?

Consumers are increasingly worried that their current consumption patterns have negative environmental impacts, which in turn shapes their green purchase intentions. Based on the signaling theory and stimulus–organism–response model, the purpose of this research is to construct a theoretical framework to understand consumer intentions to buy eco-labeled products. Empirical results from 671 questionnaires show that as expected, green advertising receptivity positively affects intention, and the relationship between green advertising receptivity and intention is also moderated by promotion focus and mediated by system trust and personal trust. However, the relationship between green advertising receptivity and purchase intention is not moderated by prevention focus. Hence, this research suggests that stakeholders should conduct truthful green advertising campaigns to dispel consumer suspicion and target different consumers with different green advertising and marketing campaigns to increase sales.     

Consumers can learn and can forget – Modeling the dynamic decision procedure when watching TV

Facing the challenge of attracting consumers and winning market share under the proliferation of TV stations and channels, the traditional TV stations often make some marketing strategies. However, how to evaluate the effectiveness of different strategies and select the best one is a key issue. This study proposes to resolve this problem. We develop an innovative structural model to simulate the dynamic choices consumers make under two interactive behaviors: learning and forgetting. Learning behavior refers to updating programme quality assessment by using experience, while forgetting behavior prevents the use of previous experience. The Bayesian rules are employed to model learning behavior, and they are extended by incorporating an exponential decay function to measure the effect of forgetting behavior. The structural model is tested and validated by using Hong Kong television viewing data. The empirical results show that when modeling consumer choice decisions, considering learning and forgetting behavior significantly improves the performance of the model in regard to rating prediction and marketing strategy evaluation. Five cases are simulated to show how the model is used to evaluate marketing strategies. Managerial implications are then discussed to guide the decision-making of traditional TV broadcasters and advertisers.     

客户终生价值测度模型的应用研究——基于大连友好商城的案例分析

本案例研究以客户终生价值测度模型为基本工具,通过挖掘客户历史购买信息,在预测客户未来购买情况的基础上对客户终生价值进行测度,探索其对企业实践的管理意义.以大连友好商城为案例,针对其当前存在的问题,探讨并选择合适的客户终生价值测度模型;通过Pareto/NBD模型预测客户未来的购买次数,Gamma-Gamma模型预测未来的平均购买金额,在此基础上计算客户终生价值;对商城现有客户进行分类,深入分析了不同类别客户的购买特征,提出相应的差异化营销策略.研究表明,通过测度客户终生价值,可以帮助企业识别并分类客户,进而实施差异化的营销策略,提升企业竞争优势.     

A Bayesian approach for predicting food and beverage sales in staff canteens and restaurants

Accurate demand forecasting is one of the key aspects for successfully managing restaurants and staff canteens. In particular, properly predicting future sales of menu items allows for a precise ordering of food stock. From an environmental point of view, this ensures a low level of pre-consumer food waste, while from the managerial point of view, this is critical to the profitability of the restaurant. Hence, we are interested in predicting future values of the daily sold quantities of given menu items. The corresponding time series show multiple strong seasonalities, trend changes, data gaps, and outliers. We propose a forecasting approach that is solely based on the data retrieved from point-of-sale systems and allows for a straightforward human interpretation. Therefore, we propose two generalized additive models for predicting future sales. In an extensive evaluation, we consider two data sets consisting of multiple time series collected at a casual restaurant and a large staff canteen and covering a period of 20 months. We show that the proposed models fit the features of the considered restaurant data. Moreover, we compare the predictive performance of our method against the performance of other well-established forecasting approaches.     

我国轿车性价比评价模型

我国轿车市场日渐成熟,消费者购车更加理性,希望购买到性价比高的车子。目前市场上关于轿车性价比的评价多为定性的描述。运用价值工程理论构建出我国家用轿车性价比的评价模型,对家用轿车进行定量描述,期望能为消费者的购车提供一定的参考。