一种基于LSTM模型的日销售额预测方法

精准的销售额预测对于商业运营有非常大的指导意义,可以指导运营后台提前进行合理的资源配置,帮助管理者制定合理的目标。零售商店日销售额预测指从商店已有日销售额的数据资料中总结出商品销售额的变化规律,并根据该规律动态预测未来一段时间内的日销售额。预测目的是通过增加企业销量,从而完善生产模式,使企业获利。目前,现有的关于商品销售额预测方法的精度大都不高,低于85%。因此,提出了一种基于TensorFlow的LSTM模型的零售商店日销售额预测方法,能够提高预测未来一周的日销售额精度。实验结果显示,预测精度达到90%;同时得到LSTM模型的MAPE为0.031932,MAE为168.3207,明显高于现有模型的预测结果。     

基于混沌果蝇支持向量机回归的产品销售预测

针对制造业产品销售时序具有多维、小样本、非线性、多峰等特征,提出一种混沌果蝇支持向量机回归的产品销售预测方法。将混沌理论引入到果蝇优化算法中,从而提高果蝇种群多样性和搜索的遍历性,并在寻优过程中加入混沌扰动,避免搜索过程陷入局部最优,增加持续搜索可行解的能力。并用算例验证了混沌果蝇优化算法（Chaos Fruit Fly Optimization Algorithm,CFOA）的优化性能,通过优化支持向量机回归（Support Vector Regression,SVR）的参数构建销售预测模型,进行了汽车零部件销售预测。结果表明基于混沌果蝇支持向量机回归的产品销售预测方法是有效可行的。     

基于多维灰色模型及神经网络的销售预测

在时尚销售领域,如服饰、手袋、钱包等,准确的销售预测对企业非常重要.然而由于客户的需求受诸多因素的影响,要做到准确的销售预测一直是一个富有挑战性的问题.基于改进的多维灰色模型（GM（1,N））和神经网络（ANN）提出一种混合模型来预测销量,其中多维灰色模型对销售数据建模,神经网络对误差进行校正.该混合模型的优点是考虑了影响客户需求的因素与销量之间的关系.通过对阿里天猫销售数据来评估混合模型的表现,实验结果表明,所提出的混合模型的预测结果要优于其他几种销售预测模型.     

基于ARMA的销售预测方法与系统实现

为了提高销售预测准确性,为企业生产决策提供参考依据,建立一个基于自回归滑动平均模型ARMA的销售预测模型,实现产品销售预测。采用修正因子对输入序列进行影响因素权值调整（前处理）,再进行ARMA建模,并对预测结果再进行修正（后处理）,提高了销售预测的准确性。以IIS为应用服务器,Oracle为数据库服务器,采用B/S体系结构和ASP.NET四层架构设计,实现时序销量数据修正、模型的识别、定阶、参数估计和预测数据修正以及预测展示等功能,完成产品销量预测系统。     

ProfARIMA: A profit-driven order identification algorithm for ARIMA models in sales forecasting

In forecasting, evolutionary algorithms are often linked to existing forecasting methods to optimize their input parameters. Traditionally, the fitness function of these search heuristics is based on an accuracy measure. In this paper, however, we combine forecasting accuracy with business expertise by defining a flexible and easily interpretable profit function for sales forecasting, which is based on the profit margin of a given product, the volume of its sales and the accuracy of the forecast. ProfARIMA is a new procedure that selects the lags of a Seasonal ARIMA model according to the profit of a model's forecasts by taking advantage of search heuristics. This procedure is tested on both publicly available datasets and a real-life application with datasets of The Coca-Cola Company in order to assess its performance, both in profit and accuracy. Three different evolutionary algorithms were implemented during this testing process, i.e. Genetic Algorithms, Particle Swarm Optimization and Simulated Annealing. The results indicate that ProfARIMA always performs at least equally to the Box–Jenkins methodology and often outperforms this traditional procedure. For The Coca-Cola Company, our new algorithm in combination with Genetic Algorithms even leads to a significantly larger profit for out-of-sample forecasts.     

Optimal pricing of online products based on customer anchoring-adjustment psychology

Online product pricing is an important factor that affects the purchase of customers and the earnings of e-commerce platform enterprises. It has an increasingly prominent influence on consumers’ purchasing decisions. To study the influence of customers’ anchoring psychology and product pricing on customers’ purchasing behaviors, this paper proposes a pricing model that considers online consumers to be anchored by the product price and sales. The cognitive bias of online consumers is described by a utility function based on an anchoring-adjustment heuristic, and consumer choice behavior and an online product pricing strategy are studied with a multinomial logit model. Then, it studies the influence of the anchoring point and anchoring degree on optimal pricing, sales and profit. In particular, when consumers are anchored only by sales of goods, the optimal solution is monotonic with respect to the anchoring degree. However, when consumers are anchored by both price and sales, both factors affect consumer behaviors and produce more complex results. This paper provides a flexible pricing mechanism for platform enterprises, and it can provide a theoretical basis and decision support for dynamic pricing of platform enterprises based on historical consumption data.     

基于交叉验证网格寻优支持向量机的产品销售预测

综合考虑影响汽车销售的多种因素,运用交叉验证网格搜索优化支持向量机的惩罚系数和核函数参数,建立了适合汽车销售的预测模型.仿真实验结果表明,改进支持向量机优化汽车销售预测模型的预测效果比某公司当前采用的模型更佳,该模型具有较高的预测精度和较大的可信度,可为企业决策层提供较为准确的销售预测参考.     

Modified diffusion model with multiple products using a hybrid GA approach

As technology advances, the speed in which new products are developed also increases. Due to such increases, product forecasting has become much more vital for a company. The Bass diffusion model is a demand-forecast model that explores the phases of a product’s life cycle that have been successful in the diffusion of forecasting innovation in new products. Recognizing the need for an efficient parameter estimation method for multi-product forecasting, we have conducted research using the hybrid genetic algorithm (HGA). The research conducted will provide an alternate approach to explore the forecasting capability of the diffusion models without having as many limitations as the original method. We used both published data and LCD-monitor global sales data to test and verify our method. Results show that the proposed model using a hybrid GA approach can improve the forecasting efficiency.     

小样本下多维指标融合的电商产品销量预测

为突破传统预测方法在小样本数据下电商产品销量预测中精度较低的局限，开展基于集成学习Xgboost的预测模型研究。综合考虑影响电商产品销量的多维指标，包括：在线搜索、在线评论、页面访问、库存与订购量、情绪指数等并利用熵值法融合同类指标。应用Logistic函数和正则修正项，结合贪心算法划分子树，构建基于集成学习Xgboost的电商产品销量预测模型。针对京东商城的联想zuk z2手机产品进行模型检验，并与BP神经网络、SVM支持向量机、BP-SVM组合预测三个模型进行对比，发现融合多维指标的Xgboost预测模型的精度显著提高，为小样本数据下电商产品销量预测提供方法和思路。     

基于商品聚类的电商销量预测

随着我国大力推进电商行业的发展,越来越多的电商企业加入到线上的竞争之中.随着销量的增大,第三方电商企业所掌握的销售数据也越来越多,这些分类上零散的销售数据给数据处理预测带来了一定的难度,常常导致在预测过程中数据不完备或者预测结果存在非常大的偏差.为了改善这一问题,这里提出了一种基于销售数据的产品重分类预测模型,利用产品销售共性提取产品聚类簇,再使用时间序列模型得出预测结果并通过隐马尔科夫预测模型给出预测结果的概率分布.通过实验分析,利用以上模型的预测获得较好的预测结果,对电商企业制定营销策略具有一定的参考价值.     

基于群体决策特征的中式自选餐厅菜品销量预测模型

从中式自选餐厅的运营流水数据中刻画餐饮消费者群体的膳食特征,并对餐厅菜品销量进行预测。由于中式自选餐厅每日供应的菜品变化巨大,传统的时间序列预测方法难以解决问题,为此提出了一种基于决策理论的预测模型。通过MACBETH方法和期望效用理论将餐厅的菜品转换为统一的效用值,作为就餐群体选择菜品的依据,使用决策权重代表消费者群体的膳食特征,然后基于行为决策理论设计餐饮消费者的决策模型来预测各个菜品的销量。使用了中式自选餐厅的流水数据训练销量预测模型,模型在训练集的交叉熵为0.082,在测试集上的交叉熵为0.086。相比基于神经网络的预测模型,该模型精确度更高,并具有良好的解释性。因此,该模型能够精确预测给定菜品供应下中式自选餐厅的菜品销量,从而支持中式自选餐厅的高效运营。     

A new forecasting scheme for evaluating long‐term prediction performances in supply chain management

Supply chain management (SCM) practitioners in inventory sites are often required to predict the future sales of products in order to meet customer demands and reduce inventory costs simultaneously. Although a variety of forecasting methods have been developed, many of them may not be used in practice for various reasons, such as insufficient viable information about sales and oversophisticated methods. In this paper, we provide a new forecasting scheme to evaluate long‐term prediction performances in SCM. Three well‐known forecasting methods for time series data—moving average (MA), autoregressive integrated MA, and smoothing spline—are considered. We also focus on two representative sales patterns, each of which is with and without a growth pattern, respectively. By applying the proposed scheme to various simulated and real datasets, this research aims to provide SCM practitioners with a general guideline for time series sales forecasting, so that they can easily understand what prediction performance measures and which forecasting method can be considered.     

Retracted: A hybrid artificial intelligence sales‐forecasting system in the convenience store industry

Recently, there has been increasing interest in computer‐aided ergonomics and its applications, such as in the fields of intelligent robots, intelligent mobiles, intelligent stores, and so on. The operation of convenience stores (CVS) in Taiwan is facing a crossover revolution by providing multiple services, including daily fresh foods, a café, ticketing, and a grocery. Therefore, forecasting the daily sales of fresh foods is getting more and more complex due to the influence of both internal and external factors. Eventually, a reliable sales‐forecasting system will play an important role in improving business strategies and increasing competitive advantages. The purpose of this study is the development of an enhanced hybrid sales‐forecasting model of fresh foods, called ECFM (Enhanced Cluster and Forecast Model), for CVSs by combining a self‐organization map (SOM) neural network and radial basis function (RBF) neural networks. The model is evaluated for a six‐month sales data set of daily fresh foods at a chained CVS in Taiwan. Meanwhile, the performance of the proposed model is compared with that of fuzzy neural network (FNN) and cluster and forecast model (CFM). The result reveals that the proposed model is not only amenable but can also promise the fresh food sales forecasting for CVSs. © 2011 Wiley Periodicals, Inc.     

An improved sales forecasting approach by the integration of genetic fuzzy systems and data clustering: Case study of printed circuit board

Success in forecasting and analyzing sales for given goods or services can mean the difference between profit and loss for an accounting period and, ultimately, the success or failure of the business itself. Therefore, reliable prediction of sales becomes a very important task. This article presents a novel sales forecasting approach by the integration of genetic fuzzy systems (GFS) and data clustering to construct a sales forecasting expert system. At first, all records of data are categorized into k clusters by using the K-means model. Then, all clusters will be fed into independent GFS models with the ability of rule base extraction and data base tuning. In order to evaluate our K-means genetic fuzzy system (KGFS) we apply it on a printed circuit board (PCB) sales forecasting problem which has been used as the case in different studies. We compare the performance of an extracted expert system with previous sales forecasting methods using mean absolute percentage error (MAPE) and root mean square error (RMSE). Experimental results show that the proposed approach outperforms the other previous approaches.     

A hybrid extreme learning machine model with harris hawks optimisation algorithm: an optimised model for product demand forecasting applications

Accurate and real-time product demand forecasting is the need of the hour in the world of supply chain management. Predicting future product demand from historical sales data is a highly non-linear problem, subject to various external and environmental factors. In this work, we propose an optimised forecasting model - an extreme learning machine (ELM) model coupled with the Harris Hawks optimisation (HHO) algorithm to forecast product demand in an e-commerce company. ELM is preferred over traditional neural networks mainly due to its fast computational speed, which allows efficient demand forecasting in real-time. Our ELM-HHO model performed significantly better than ARIMA models that are commonly used in industries to forecast product demand. The performance of the proposed ELM-HHO model was also compared with traditional ELM, ELM auto-tuned using Bayesian Optimisation (ELM-BO), Gated Recurrent Unit (GRU) based recurrent neural network and Long Short Term Memory (LSTM) recurrent neural network models. Different performance metrics, i.e., Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE) and Mean Percentage Error (MPE) were used for the comparison of the selected models. Horizon forecasting at 3 days and 7 days ahead was also performed using the proposed approach. The results revealed that the proposed approach is superior to traditional product demand forecasting models in terms of prediction accuracy and it can be applied in real-time to predict future product demand based on the previous week’s sales data. In particular, considering RMSE of forecasting, the proposed ELM-HHO model performed 62.73% better than the statistical ARIMA(7,1,0) model, 40.73% better than the neural network based GRU model, 34.05% better than the neural network based LSTM model, 27.16% better than the traditional non-optimised ELM model with 100 hidden nodes and 11.63% better than the ELM-BO model in forecasting product demand for future 3 months. The novelty of the proposed approach lies in the way the fast computational speed of ELMs has been combined with the accuracy gained by tuning hyperparameters using HHO. An increased number of hyperparameters has been optimised in our methodology compared to available models. The majority of approaches to improve the accuracy of ELM so far have only focused on tuning the weights and the biases of the hidden layer. In our hybrid model, we tune the number of hidden nodes, the number of input time lags and even the type of activation function used in the hidden layer in addition to tuning the weights and the biases. This has resulted in a significant increase in accuracy over previous methods. Our work presents an original way of performing product demand forecasting in real-time in industry with highly accurate results which are much better than pre-existing demand forecasting models.

     

基于聚类与机器学习的零售商品销量预测

本文提出一种基于K-means聚类与机器学习回归算法的预测模型以解决零售行业多个商品的销售预测问题,首先通过聚类分析识别出具有相似销售模式的商品从而实现数据集的划分,然后分别在每个子数据集上训练了支持向量回归、随机森林以及XGBoost模型,通过构建数据池的方式增加了用于训练模型的数据量以及预测变量的选择范围.在一家零售企业的真实销售数据集上对提出的模型进行了验证,实验结果表明基于K-means和支持向量回归的预测模型表现最优,且所提出的模型预测效果明显优于基准模型以及不使用聚类的机器学习模型.     

Impact of online word-of-mouth on sales: the moderating role of product review quality

Facing with thousands of online product reviews, consumers usually pay close attention to those valuable ones which provide more specific and credible evaluations on products. Whether a close association exists between product review quality and sales is thus examined in this paper. By employing text mining techniques on multiple review features, a review is measured as one of the following two levels: high-quality or low-quality. In doing so, aggregate quality level of product’s whole reviews is also identified. Then, a two-level econometrical analysis is conducted on the real datasets from Amazon.cn. The results reveal that aggregate quality level of positive reviews and negative reviews interactively influence sales. In the situation the aggregate quality level of positive reviews is high meanwhile that of negative reviews’ is low, product sale is the highest, while in the opposite situation product sale is the lowest. The results also reveal that consumers understand product’s value from weighting positive and negative reviews of high-quality level, which then positively relates to product sales and exerts a dynamic effect on sales by the moderating role of product selling stage and popularity. The paper innovatively integrates the quantitative and qualitative characteristics of reviews to estimate their economic effect.     

Is online multiple-stores cooperative promotion better than single-store promotion? Misprediction from evaluation mode

Inspired by the evaluation mode theory, we show online vendors mispredict consumers’ responses to different types of sales displays. While vendors predict that consumers evaluate a featured product more positively in a cooperative sales promotion (CSP; i.e., multiple stores promoting synchronously) than in an independent sales promotion (ISP; i.e., a single store promoting independently), consumers actually do the opposite. The reason is that vendors compare CSP with ISP and are able to evaluate depending on perceived acquisition utility; consumers, however, see either CSP or ISP, resulting in difficulty in accessing acquisition utility. Thus, they evaluate according to perceived transaction utility.     

企业品牌形象与产品造型设计

品牌形象是企业最有价值的资产之一.现在许多企业已经意识到良好的品牌形象能够使他们更加快速有效的发展.良好的产品造型设计是吸引消费者最直接的途径,也是消费者购买动机的主要因素之一.对一个完整的产品设计来说,不仅要规范生产,销售和产品本身,还需要注意到不同品牌形象产品造型的不同需求,独特而良好的造型风格能够增强产品的可识别性.树立良好的品牌形象,从而提高消费者的忠诚度.     

A hybrid SARIMA wavelet transform method for sales forecasting

Time series forecasting, as an important tool in many decision support systems, has been extensively studied and applied for sales forecasting over the past few decades. There are many well-established and widely-adopted forecasting methods such as linear extrapolation and SARIMA. However, their performance is far from perfect and it is especially true when the sales pattern is highly volatile. In this paper, we propose a hybrid forecasting scheme which combines the classic SARIMA method and wavelet transform (SW). We compare the performance of SW with (i) pure SARIMA, (ii) a forecasting scheme based on linear extrapolation with seasonal adjustment (CSD + LESA), and (iii) evolutionary neural networks (ENN). We illustrate the significance of SW and establish the conditions that SW outperforms pure SARIMA and CSD + LESA. We further study the time series features which influence the forecasting accuracy, and we propose a method for conducting sales forecasting based on the features of the given sales time series. Experiments are conducted by using real sales data, hypothetical data, and publicly available data sets. We believe that the proposed hybrid method is highly applicable for forecasting sales in the industry.