Clothing Sales Forecast Considering Weather Information: An Empirical Study in Brick-and-Mortar Stores by Machine-Learning

Reliable sales forecasts are important to the garment industry. In recent years, the global climate is warming, the weather changes frequently, and clothing sales are affected by weather fluctuations. The purpose of this study is to investigate whether weather data can improve the accuracy of product sales and to establish a corresponding clothing sales forecasting model. This model uses the basic attributes of clothing product data, historical sales data, and weather data. It is based on a random forest, XGB, and GBDT adopting a stacking strategy. We found that weather information is not useful for basic clothing sales forecasts, but it did improve the accuracy of seasonal clothing sales forecasts. The MSE of the dresses, down jackets, and shirts are reduced by 86.03%, 80.14%, and 41.49% on average. In addition, we found that the stacking strategy model outperformed the voting strategy model, with an average MSE reduction of 49.28%. Clothing managers can use this model to forecast their sales when they make sales plans based on weather information.

Parameters Optimization Using Genetic Algorithms in Support Vector Regression for Sales Volume Forecasting

Budgeting planning plays an important role in coordinating activities in organizations. An accurate sales volume forecasting is the key to the entire budgeting process. All of the other parts of the master budget are dependent on the sales volume forecasting in some way. If the sales volume forecasting is sloppily done, then the rest of the budgeting process is largely a waste of time. Therefore, the sales volume forecasting process is a critical one for most businesses, and also a difficult area of management. Most of researches and companies use the statistical methods, regression analysis, or sophisticated computer simulations to analyze the sales volume forecasting. Recently, various prediction Artificial Intelligent (AI) techniques have been proposed in forecasting. Support Vector Regression (SVR) has been applied successfully to solve problems in numerous fields and proved to be a better prediction model. However, the select of appropriate SVR parameters is difficult. Therefore, to improve the accuracy of SVR, a hybrid intelligent support system based on evolutionary computation to solve the difficulties involved with the parameters selection is presented in this research. Genetic Algorithms (GAs) are used to optimize free parameters of SVR. The experimental results indicate that GA-SVR can achieve better forecasting accuracy and performance than traditional SVR and artificial neural network (ANN) prediction models in sales volume forecasting.

A New Type of Combination Forecasting Method Based on PLS——The Application of It in Cigarette Sales Forecasting

Cigarette market is a kind of monopoly market which is closed loop running, it depends on the plan mechanism to schedule producing, supplying and selling, but the “bullwhip effect” still exists. So it has a fundamental significance to do sales forecasting work. It needs to considerate the double trend characteristics, history sales data and other main factors that affect cigarette sales. This paper depends on the panel data of A province’s cigarette sales, first we established three single forecasting models, after getting the predicted value of these single models, then using the combination forecasting method which based on PLS to predict the province’s cigarette sales of the next year. The results show that the prediction accuracy is good, which could provide a certain reference to cigarette sales forecasting in A province.

Consumers'Purchase Intention of Geographical Indication Agricultural Products under the Background of Live-streaming Sales:An Exploratory Research Based on Grounded Theory

Exploring the mechanism for the formation of consumer purchase intentions of geographical indication agricultural products in the context of live-streaming sales can provide an important reference for brand marketing of geographical indication agricultural products.In this study,in-depth interviews were conducted with consumers of geographical indication agricultural products.Based on grounded theory,open coding,axial coding and selective coding were performed for interview text.Finally,21 concepts,7 subcategories and 3 main categories were obtained,and a model of the formation mechanism of the purchase intention of geographical indication agricultural products under the background of live-streaming sales was constructed,that is,"consumer cognition-consumer attitude-consumer behavior".Among them,consumer cognition includes two dimensions:the type of geographical indication agricultural products and the live-streaming appeal strategy,i.e.,the personal cognition of consumer and the promotion of live-streaming host's strategy.Consumer attitude is value perception of consumers,mainly including two dimensions of functional value and emotional value.Consumer behavior is the consumer's willingness to buy.It has been concluded that the types of geographical indication agricultural products interact with the live-streaming appeal strategies.Through the intermediary of consumers'value perception,consumers'purchase intention is generated.Among them,resource-oriented geographical indication agricultural products adopt rational live-streaming appeal strategies,which can enhance the consumer's perception of functional value,thereby promoting their purchase intention;and cultural and creative geographical indication agricultural products brands adopt perceptual live-streaming appeal strategies,which can enhance the emotional value perception of consumers,thereby promoting their purchase intention.

Comparison of Sales Prediction in Conventional Insights and Machine Learning Perspective

With the integration of global economy development and the rapid growth of science knowledge and technology,the needs of people’s consumption are increasingly personalized and diversified.Such a market background makes sales forecasting become an indispensable part of enterprise management and development.The definition of the sales forecasting is that based on the past few years’sales situation,the enterprises through systematic sales forecasting models estimate of the quantity and amount of all or some specific sales products and services in a specific time in the future.Accurate sales forecasting can promote enterprises to do better in future revenue,and can also encourage enterprises to set and keep an efficient sales management team.This paper will analyze traditional sales forecasting methods and sales forecasting methods based on big data models related to the perspective of machine learning,and then compare them.The research shows that the two sales forecasting methods have their own advantages and disadvantages.In the future,enterprises can adopt the two sales forecasting methods in parallel to maximize the utilization advantage of sales forecasting for enterprises.

A Type of Combination Forecasting Method Based on Time Series Method and PLS

This paper depends on the panel data of Anhui province and its 17 cities’ cigarette sales. First we established three single forecasting models (Holter-Wintel Season product model, Time series model decomposing model and Partial least square regression model), after getting the predicted value of cigarette sales from these single models, we then employ the combination forecasting method based on Time Series method and PLS to predict the province and its 17 cities’ cigarette sales of the next year. The results show that the accuracy of prediction is good which could provide a reliable reference to cigarette sales forecasting in Anhui province and its 17 cities.

消费行为数据采集平台的安全保障与预测模型研究

依据用户浏览记录等信息进行兴趣爱好的预测并进行合理推荐,已成为诸多销售平台优化用户体验的常用手段,而用户信息安全问题自然也成了各大平台面临的一大挑战.提出一种基于内生安全的消费行为数据采集与分析平台,通过采集用户数据,使用基于长短时记忆网络的预测模型,精准预测未来销售流量数据.在数据安全性方面,平台使用基于内生安全的拟态云WAF,通过动态选择算法、异构执行体和裁决算法3种核心技术为整个数据平台提供了自主可控的安全保障,并利用基于Sketch的网络测量技术对异常流量进行了检测.此外,平台融合了数据备份和恢复、加密存储、数据传输加密技术,并对重要的数据采取分类存储、访问控制等措施.多项对比实验验证表明,用于中烟销售流量的预测平台相较于目前提出的多种技术在预测准确度和数据安全方面都有显著提升,可为企业销量预测提供一种合理可行的解决方案.

基于卷积神经网络的电力市场短期售电量预测方法

电力市场短期售电量预测的精度对优化用电结构以及提高供电可靠性具有重要意义,传统短期售电量预测方法没有考虑偏差电量考核影响、用电行为差异导致电预测精度低,提出基于卷积神经网络的电力市场短期售电量预测方法,首先根据用户的用电负荷率进行分类,获取不同行业的用电特征和需求模式,然后考虑正负偏差电量的影响,设计基于CNN-ResNet的短期售电量预测方法,通过实验分析表明,该方法能够有效提高多因素影响下售电量预测的准确率。

基于Bass模型的新能源汽车销量及碳减排量预测

基于Bass扩散模型,预测我国2023-2030年新能源汽车销量,考虑纯电动汽车和混合动力汽车的碳减排效果差异,计算纯电动汽车和混合动力汽车的平均市场份额,从而测算我国2023-2030年新能源汽车全生命周期碳减排量。结果表明,在考虑能源结构等因素变化的情景下,相较于传统汽油乘用车,推广使用新能源汽车,到2030年碳减排量预计可达4.31亿t,预计减排34.73%。

线上销售能否改善资本市场信息环境——基于分析师盈余预测的证据

基于2015—2019年上市公司通过互联网平台进行线上销售的独特数据,本文研究了线上销售对分析师盈余预测的影响。研究发现:(1)上市公司进行线上销售显著降低了分析师盈余预测误差和分歧度,提升了分析师盈余预测质量,且上述结果在仅保留线上销售行业和PSM匹配后的样本、采用Heckman两阶段模型等方式缓解内生性问题后仍然成立;(2)线上销售对分析师盈余预测质量的提升作用在分析师公开获取信息渠道较少、获取信息成本较高时更为显著;(3)机制检验发现,线上销售可以通过降低盈余波动,或者抑制盈余管理以改善公司盈余信息质量,进而降低分析师盈余预测误差和分歧度。研究结果表明,线上销售不仅增加了分析师获取公司特质信息的来源,而且一定程度上改善了盈余信息质量,有助于分析师发挥资本市场信息中介的重要作用,提高资本市场效率。

基于层次化多模态注意力机制循环神经网络的服装新品销售预测

在新服装产品销售预测任务中,由于缺乏历史销售数据,通常需要充分利用其他模态的数据作为补充。然而,多模态服装数据通常具有冗余性和异构性。为解决这些问题,提出一种包括三个主要元素的层次化多模态注意力循环神经网络(hierarchical multi-modal attention based recurrent neural network,HMA-RNN)。层次化结构将高层语义信息与低层语义信息分离,以避免信息冗余。在模态融合阶段引入多模态注意力机制(multi-modal attention,MMA)以减轻固有的数据不对齐问题。采用共享注意力机制构建跨多模态数据的依赖关系。在Visuelle 2.0数据集上的试验结果表明,所提出的方法加权平均百分比误差(weighted average percentage error,WAPE)为72.07,平均绝对误差(mean absolute error,MAE)为0.80,明显优于现有的方法,表明了该研究所提出的方法的有效性。

基于改进注意力机制Transformer网络的快消品销量预测方法

销量预测能为企业生产计划、仓储运输提供决策支持,使企业能更好地适应市场需求。快消品销售量受众多因素的影响,具有季节性和周期性规律,传统的线性模型难以准确的预测,本文从长时序列预测的视角,运用深度学习理论,提出了一种基于订单时序和订单频率的改进自注意力机制模型(Sequence-Frequency Transformer,SFTransformer)。首先,基于快消品订单数据构建原始数据集,采用time2vec编码处理订单时序信息,并融合订单数据的时序和频率特征在基于时序的订单数据的不同订单频率分别对应不同的注意力头来关注订单数据的订单时序特征和频率特征;使用Transformer模型架构提取特征进行长时序列预测。在真实数据集上进行对比实验,SFTransformer模型在均方误差(MSE)、平均绝对误差(MAE)、均方根误差(RMSE)3项指标上均取得了最佳性能,验证了本文所提方法的有效性。

基于GWO-Prophet的商品销售预测研究

零售企业的各项经营活动都离不开商品的销售情况,对商品的销售预测为企业制定生产计划与经营决策等活动提供重要的依据。针对企业销售额预测中销售额时间序列受外界条件影响大,预测精度低等问题,论文提出了一种基于GWO-Prophet的商品销售预测方法。基于某零售企业2015年-2018年销售额数据,通过Prophet模型将高维的销售额数据分别构建对应趋势项、季节项、节假日项、残差项的低维时序特征分量,分别用这些低维特征分量进行拟合后通过加法模型累加来预测未来一年的销售额数据;通过灰狼寻优算法(GWO)对Prophet模型参数进行智能寻优,防止模型陷入局部最优从而提高模型的精确度,通过灰狼寻优算法优化后的Prophet模型能更好地拟合突变点,季节项,节假日项等外界因素对销售额的影响。以MAE、MAPE和RMSE作为模型评估的指标,结果表明,基于GWO-Prophet模型的预测精度不仅优于单一的Prophet模型,还优于其他如ARIMA、SARIMA、LSTM对比模型。

基于Stacking集成学习的商品销量预测研究

销量预测准确度的提升能够帮助电商企业更有效地规划库存,从而提高供应链管理的效率。文中利用Stacking集成学习方法融合多模型来预测销量,研究通过对基学习器进行选择,以RF、SVM、XGBoost、LSTM算法为第一层模型,线性回归为第二层模型,通过历史销售数据集进行验证。研究发现,相较于单一模型,Stacking集成学习方法的预测效果更优。此研究方法可以有效融合多种模型优势,得到更准确的销量预测结果,从而为电商企业制定库存和生产计划提供依据。

基于销量预测的LSTM模型优化

为了解决销量预测精度不准确以及数据集贫乏的问题,实现结果更精准、效率更高的销量预测,提出一种基于特征工程的长短时记忆(LSTM)神经网络销量预测模型.LSTM模型是经过循环神经网络(RNN)模型优化后的一种特殊结构,适用于处理与时间序列相关的问题.通过特征工程提取、筛选、衍生出新的特征序列与原始特征序列融合,并结合LSTM算法进行销量预测,提高了LSTM模型预测的精准度.以Rossmann店铺的实际销售数据为例,证实了经过优化后的特征序列集融合LSTM模型进行预测,在提高预测效率的同时也提升了预测结果的精准度.

基于VMD-CNN-GRU-LSTM组合模型的汽车销量预测分析

针对汽车销量时间序列数据的季节性、非线性性、非平稳性等复杂特征,提出一种融合变分模态分解(Variational Mode Decomposition,VMD)的卷积神经网络(Convolutional Neural Network,CNN)、门控循环单元(Gated Recurrent Unit,GRU)和长短期记忆(Long Short-Term Memory,LSTM)神经网络组合的汽车销量预测方法,通过VMD将汽车销量时序数据进行分解,利用CNN提取关键特征,并通过GRU与LSTM捕捉汽车时序数据的时间依赖关系.实验表明该方法有较好的预测性能.

基于电商数据的皮革服装销量预测模型构建及仿真

为进一步提高皮革服装的销售预测精度,结合传统BP神经网络在自适应和学习方面的优势,提出一种模拟退火算法改进BP神经网络的皮革服装销量预测模型。结果表明,模拟退火算法优化后的BP神经网络只需迭代14次即达到稳定,同时优化后的BP神经网络的皮革服装销量预测精度要明显高于传统的BP神经网络。由此可以证明,本文改进的BP神经网络具有一定的优势。

考虑多影响因素的加油站销量预测研究

加油站销量预测是成品油二次物流中不可或缺的一部分,有利于推进油品主动配送,帮助二次物流运输精确调度,提升企业整体运行效率。通过历史数据及模型结果分析得出,加油站销量与特殊日期、长短时等多因素有关,文章在长短期记忆网络(LSTM)基础上引入多因素特征构建预测模型,并以某市加油站真实销量数据为实例进行检验。经验证,模型在汽油预测方面的平均误差小于15%,柴油预测的平均误差小于18%,能够满足应用需求。

采用节能导线的输电线路全寿命周期成本分析

近20年来,全国输电线路的长度显著增长,在线路工程建设中,经济性成本是至关重要的考量因素之一。通过构建输电线路工程全寿命周期成本量化模型,对辽宁省A市某220 kV新建线路工程采用JL/G1A-400/35型钢芯铝绞线及JL(GD)/G1A-400/35型钢芯高导电率铝绞线下的经济成本差异进行对比分析。采用线性回归分析、灰色预测GM(1,1)模型以及自回归差分移动平均(autoregressive integrated moving average,ARIMA)模型对工程运行期输电线路的负荷水平进行预测,得出组合预测分析方法下线路工程的售电效益;同时,量化节能导线的减碳效益,分析负荷增长水平、天气、海拔及投资折现率对线路运行成本的灵敏度,对比评估2种导线技术方案的经济性指标。研究结果表明,作为一种节能导线,钢芯高导电率铝绞线可以降低输电线路工程的运行维护成本,在考虑售电效益的同时为线路工程带来额外的减碳效益,在工程的全寿命周期内具有良好的经济性及低碳性。

基于机器学习算法的服装直播销量预测模型

为了完善直播销量预测的指标体系,提高直播销量预测的准确率,本文对比多种机器学习算法,分析服装直播销量的影响因素并预测服装直播销量。首先,通过文献分析选取直播销量的影响因素,并使用Spearman相关系数结合显著性进行筛选。其次,通过不同机器学习算法建立服装销量预测模型。选用R 2、MAE、RMSE及MAPE为评价指标,采用5折交叉验证的方法,测试各模型性能。研究结果表明:主播粉丝数、主播近30天场均观看人次、主播近30天场均坑产、产品价格、产品讲解时长、产品近30天历史销量、品牌粉丝数、品牌近30天历史销量、折扣这9个因素之间共线性较弱且与直播销量之间的相关性显著,可作为预测模型中的影响因素;预测算法中K近邻算法和随机森林算法的表现较好,R 2均大于0.98,MAPE均在30.5%以内。预测结果可帮助零售商规划库存,调整生产计划,为产品采购、定价、推广提供数据支持。