Search

Although live streaming is indispensable, live-streaming e-business requires accurate and timely sales-volume prediction to ensure a healthy supply–demand balance for companies. Practically, because various factors can significantly impact sales results, the development of a powerful, interpretable model is crucial for accurate sales prediction. In this study, we propose SaleNet, a deep-learning model designed for sales-volume prediction. Our model achieved correct prediction results on our private, real operating data. The mean absolute percentage error (MAPE) of our model's performance fell as low as 11.47% for a + 1.5-days forecast. Even for a 1-week forecast (+ 6 days), the MAPE was only 19.79%, meeting actual business needs and practical requirements. Notably, our model demonstrated robust interpretability, as evidenced by the feature contribution results which are consistent with prevailing research findings and industry expertise. Our findings provided a theoretical foundation for predicting shopping behavior in live-broadcast e-commerce and offered valuable insights for designing live-broadcast content and optimizing the user experience. [ABSTRACT FROM AUTHOR]

Agriculture, especially cereals, is important in sustaining economies and food security globally. This study delves into the Moroccan agricultural landscape, specifically focusing on predicting livestock feed sales to assist cereal company producers in optimizing production, streamlining supply chain operations, and enhancing customer satisfaction. Data collected from various markets across Morocco, including sales dates and locations, was combined with climate data and analyzed using advanced machine learning techniques, particularly the gradient boosting regression (GBR) algorithm, which achieved high accuracy with a mean absolute error (MAE) of 0.0203 and a root mean square error (RMSE) of 0.0281. The evaluation of multiple regression models revealed promising results, demonstrating the effectiveness of predictive models in accurately forecasting sales. These findings contribute valuable insights to sales forecasting in the cereal industry by considering weather conditions, production methods, and livestock-related variables, highlighting the importance of leveraging advanced machine learning techniques for optimizing production processes and meeting market demands efficiently in the agribusiness sector. [ABSTRACT FROM AUTHOR]

Abstract Although live streaming is indispensable, live-streaming e-business requires accurate and timely sales-volume prediction to ensure a healthy supply–demand balance for companies. Practically, because various factors can significantly impact sales results, the development of a powerful, interpretable model is crucial for accurate sales prediction. In this study, we propose SaleNet, a deep-learning model designed for sales-volume prediction. Our model achieved correct prediction results on our private, real operating data. The mean absolute percentage error (MAPE) of our model’s performance fell as low as 11.47% for a + 1.5-days forecast. Even for a 1-week forecast (+ 6 days), the MAPE was only 19.79%, meeting actual business needs and practical requirements. Notably, our model demonstrated robust interpretability, as evidenced by the feature contribution results which are consistent with prevailing research findings and industry expertise. Our findings provided a theoretical foundation for predicting shopping behavior in live-broadcast e-commerce and offered valuable insights for designing live-broadcast content and optimizing the user experience.

The intricate process of planning production, involving product life cycle management and the synthesis of manufacturing information, is crucial for coherence in manufacturing. Manufacturing companies, operating in a high-mix, low-volume production environment, integrate production planning with management to focus on production processes, emphasizing high-quality, rapid product delivery. This includes material item planning to anticipate future demands and ensure sufficient raw material and finished product quantities, considering purchasing, production, and sales capacities. This study explores the electro technical sector, specifically a manufacturing entity specializing in low-voltage plastic cable distribution boxes. It scrutinizes the vital role of seasonal data cleaning in optimizing production planning, with a targeted focus on three products. The implementation of a chase demand strategy is related to capacity planning, taking into account the change in production capacity linked to demand over time. The problem in implementing this strategy is related to the fluctuating level of quality due to changes in demand for specified products.

The intricate process of planning production, involving product life cycle management and the synthesis of manufacturing information, is crucial for coherence in manufacturing. Manufacturing companies, operating in a high-mix, low-volume production environment, integrate production planning with management to focus on production processes, emphasizing high-quality, rapid product delivery. This includes material item planning to anticipate future demands and ensure sufficient raw material and finished product quantities, considering purchasing, production, and sales capacities. This study explores the electro technical sector, specifically a manufacturing entity specializing in low-voltage plastic cable distribution boxes. It scrutinizes the vital role of seasonal data cleaning in optimizing production planning, with a targeted focus on three products. The implementation of a chase demand strategy is related to capacity planning, taking into account the change in production capacity linked to demand over time. The problem in implementing this strategy is related to the fluctuating level of quality due to changes in demand for specified products. [ABSTRACT FROM AUTHOR]

Copyright of Wool Textile Journal is the property of National Wool Textile Science & Technology Information Center and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Effective sales prediction for e-commerce would assist retailers in developing accurate production and inventory control plans, which would further help them to reduce inventory costs and overdue losses. This paper develops a systematic method for e-commerce sales prediction, with a particular focus on predicting the sales of products with short shelf lives. The short-shelf-life product sales prediction problem is poorly addressed in the existing literature. Unlike products with long shelf lives, short-shelf-life products such as fresh milk exhibit significant fluctuations in sales volume and incur high inventory costs. Therefore, accurate prediction is crucial for short-shelf-life products. To solve these issues, a stacking method for prediction is developed based on the integration of GRU and LightGBM. The proposed method not only inherits the ability of the GRU model to capture timing features accurately but also acquires the ability of LightGBM to solve multivariable problems. A case study is applied to examine the accuracy and efficiency of the GRU-LightGBM model. Comparisons among other sales prediction methods such as ARIMA and SVR are also presented. The comparative results show that the GRU-LightGBM model is able to predict the sales of short-shelf-life products with higher accuracy and efficiency. The selected features of the GRU-LightGBM model are also useful due to their interpretability while developing sales strategies. [ABSTRACT FROM AUTHOR]

This paper presents a study focused on the analysis of phytosanitary treatment sales in the Souss Massa region of Morocco. The objective of the study is to predict the sales of agricultural products, particularly crop protection solutions, aiming to optimize supply chain operations and meet customer demand effectively. Data for this study are collected from multiple sources, including the Enterprise Resource Planning (ERP) system called Microsoft Dynamics AXAPTA used by a leading agricultural company operating in the region. Information such as the date of sale, farming type, climate, and specific sales locations within the Sous Massa region is gathered. Machine learning techniques are applied for forecasting. Various regression models, including the Gradient Boosting Regressor algorithm, are employed to determine the most accurate predictor. Evaluation of the models reveals promising results, with a Mean Absolute Error (MAE) of 0.0035 and a Root Mean Square Error (RMSE) of 0.0066. The results obtained by applying various regression models, including the Gradient Boosting Regressor algorithm, demonstrate promising prediction scores. These findings contribute to the field of sales prediction in the agricultural industry while considering the impact of climate conditions, farming practices, and regional factors.

This paper aims to optimize neural network algorithms in order to improve their predictive performance and meet the demand for Douyin live sales forecasting. The Douyin live data of Florasis between January 2022 and June 2022 were collected from Huitun data and preprocessed. Using the Pearson correlation coefficient, six factors that are highly correlated with live sales were selected for subsequent prediction. This paper briefly introduces the back-propagation neural network (BPNN) algorithm and analyzes its parameter optimization methods, including particle swarm optimization (PSO), the artificial bee colony (ABC) algorithm, and the beetle antler search (BAS) algorithm. Then, an improved beetle antennae search (IBAS) algorithm was developed by introducing inertia weight and used to construct an IBAS-BPNN model for predicting sales volume in Douyin live streaming. The results showed that compared with the BPNN, PSO-BPNN, and ABC-BPNN algorithms, the IBAS-BPNN algorithm had better prediction performance, with a root-mean-square error of 335.6694, a mean absolute percentage error of 0.0532%, an equilibrium coefficient of 0.9889, and a shorter training time of 90.07 s. The experimental results demonstrate the reliability of the IBAS-BPNN algorithm for predicting sales volume in Douyin live streaming, providing new insights into parameter optimization of BPNN and offering references for further research on BPNN parameter optimization. It also provides an effective method with both timeliness and high accuracy for predicting sales volume in Douyin live streaming in practical applications. Doi: 10.28991/HIJ-2023-04-02-09 Full Text: PDF

Business intelligence makes good sale prediction crucial in any commercial activity as it has a significant impact on production and supply plan. However, practical commercial data presents explicit constraints, that how to get the optimal forecasts of commodity sales under the constraints is a vital problem many researchers face. The present research proposes a prediction model which combines graph convolution neural network and node bipartite graph. Firstly, the node bipartite graph algorithm is used to merge the constraint graph and the store graph, obtaining the “store-constraint bipartite graph”. Secondly, a graph convolutional neural network integrating GRU and AR is utilized to extract temporal features (X). Finally, a fully connected network is applied to predict the optimal solution (Y) after constraint optimization. The former can effectively learn complex features of stores, meanwhile, the later combines the constraint conditions with the store, which can effectively predict the sales of goods under the constraint conditions. In terms of model performance, we compared the proposed model with the classical method such as SVR, LSTM, ARIMA. RMSE, MSE, MAE and MAPE are used for evaluation indexes, and the results show that MAPE for one month’s sales of some product from both datasets is 7.75%.

Effective sales prediction for e-commerce would assist retailers in developing accurate production and inventory control plans, which would further help them to reduce inventory costs and overdue losses. This paper develops a systematic method for e-commerce sales prediction, with a particular focus on predicting the sales of products with short shelf lives. The short-shelf-life product sales prediction problem is poorly addressed in the existing literature. Unlike products with long shelf lives, short-shelf-life products such as fresh milk exhibit significant fluctuations in sales volume and incur high inventory costs. Therefore, accurate prediction is crucial for short-shelf-life products. To solve these issues, a stacking method for prediction is developed based on the integration of GRU and LightGBM. The proposed method not only inherits the ability of the GRU model to capture timing features accurately but also acquires the ability of LightGBM to solve multivariable problems. A case study is applied to examine the accuracy and efficiency of the GRU-LightGBM model. Comparisons among other sales prediction methods such as ARIMA and SVR are also presented. The comparative results show that the GRU-LightGBM model is able to predict the sales of short-shelf-life products with higher accuracy and efficiency. The selected features of the GRU-LightGBM model are also useful due to their interpretability while developing sales strategies.

Although numerous studies have been conducted to validate the influences of reviews and ratings on the box office, there are still some important unsettled concerns. On one hand, the evaluation of reviews should take into account different aspects of the product. On the other hand, certain movies often cause fierce debates in word-of-mouth marketing, implying that involving variance will provide a deeper perspective on sales, and there is scant research on variance compared to the volume and valence of reviews. In this study, we mainly consider two aspects in the reviews: cast and plot. The cast mainly refers to the actors and directors, which means the buyer already has an inherent bias and subjective impressions, while the plot refers to the content and storyline of the movie itself, which have lower perception and experience before consumption. We further study the influence of the variance of these two aspects on the box office as well as their changing trends. In addition, we also discuss the moderating effect of these two variances on the review valence and sales. We find that (1) the variance of both attributes has a positive impact on the box office; (2) The impact of cast on the box office gradually decreases over time, while that of plot exhibits an opposite trend; (3) The final box office is affected by the valence and variance of the cast attribute, and the variance has a negative moderating effect between the valence and sales over time. [ABSTRACT FROM AUTHOR]

Accurate and efficient prediction of outbound volume of product oil is the source and basis of scientific management for product oil inventory. In order to increase the inventory efficiency and lower the inventory cost of product oil, 166 product oil depots of a company in the western region were studied to analyze the influencing factors of inventory management on the base of the related research results of inventory management and sales prediction. Thereby, the multi-stage prediction algorithm of dual-time granularity was designed, and the algorithm was integrated with the time series model algorithm library of multi-model as well as Mamdani's fuzzy inference systems to predict the monthly and daily outbound volume of product oil and conduct the quantitative analysis. As indicated by the results, the proposed algorithm and the algorithm designed by the model library could automatically select the appropriate model according to the data characteristics of the outbound volume, as well as match and predict the outbound volume of the oil depot on large scale with high quality in short time. Generally, the median of the average absolute percentage error of the prediction results is higher than 85%, the prediction confidence is close to 95%, and the average accuracy of the monthly outbound volume prediction of the application cases can reach 90%. Conclusively, the research results could provide scientific recommendations for the decision-making on inventory management of oil depot, and have practical significance for establishing a scientific and efficient modern oil supply and logistics system.

Meeting customer needs in a timely manner has a significant impact on customer satisfaction. For this reason, the planning process has successfully influenced the success of sales activities. The crucial point for the success of the planning process depends on the sales forecasts. Sales forecasting estimates the quantity required by the customer needs. It helps in determining sales targets as campaigns, pricing, brand and product communication, and distribution channels are incorporated in the sales forecast. In this paper, we use regression and artificial neural networks to predict automobile sales in Turkey. The performance of regression is compared with that of an artificial neural network, and it is shown which network is able to predict. Thus, the result of the study, automobile sales in Turkey, was predicted and compared with the actual sales for 2020. The result is that the best prediction method will determine the automobile sales in Turkey. [ABSTRACT FROM AUTHOR]

Aim: The aim of this paper is the efficient prediction of sales during festival times in an electronic showroom using a Deep Belief Network compared to AlexNet with improved accuracy. Materials and Methods: Deep Belief Network (N=10) and AlexNet algorithm (N=10) is the iteration for different times in predicting the accuracy percentage for accidents that happened. Two sample groups are considered and tested, G-power is a calculation that contains two different groups, alpha (0.05), and power (80%). Results: It was observed that the Deep Belief Network algorithm obtains an accuracy of 83.63% and the Novel Deep Belief Network has 74.12%. This NDBN appears to have a better significance of p=0.035 than the ResNet, that is p<0.05 using independent T-test analysis. Conclusion: The result proves that the Novel Deep Belief Network approaches to predicting the best retail sales store prediction have higher accuracy than the AlexNet algorithm. [ABSTRACT FROM AUTHOR]

Covid-19 has not been defeated make the economy unstable. The government increases purchasing power with a stimulus. The government's stimulus for PPnBM 0% of cars is a breath of fresh air. Sales rate increases. Even so, the ups and downs keep happening. Sales are unstable. In one to three months, five cars can be sold. It could also take 3 to 6 months, only 1 unit. Must make a strategy to avoid overcrowding of units. Like PT. Suka Fajar Ltd Medan. Companies engaged in the sale of Mitsubishi cars. As operating costs increase, management limits the acceptance of car shipments. This policy has fatal consequences. When consumers want to buy a particular car unit, it is not available. Incidents like this happened more than five times. The company's image is not good. Researched to design and build applications that can make predictions—collaborated with the linear regression. To be directed and sequential, the waterfall is used. Ensuring the application is suitable for testing with the black box. Research leads to the conclusion that the application is designed and built according to what is needed.

Intelligent vehicles refer to a new generation of vehicles with automatic driving functions that is gradually becoming an intelligent mobile space and application terminal by carrying advanced sensors and other devices and using new technologies, such as artificial intelligence. Firstly, the traditional autoregressive intelligent vehicle sales prediction model based on historical sales is established. Secondly, the public opinion data and online search index data are selected to establish a sales prediction model based on online public opinion and online search index. Then, we consider the influence of KOL (Key Opinion Leader), a sales prediction model based on KOL online public opinion andonline search index is established. Finally, the model is further optimized by using the deep learning algorithm LSTM (Long Short-Term Memory network), and the LSTM sales prediction model based on KOL online public opinion and online search index is established. The results show that the consideration of the online public opinion and search index can improve the prediction accuracy of intelligent vehicle sales, and the public opinion of KOL plays a greater role in improving the prediction accuracy of sales than that of the general public. Deep learning algorithms can further improve the prediction accuracy of intelligent vehicle sales. [ABSTRACT FROM AUTHOR]

Microblogs are one of the major social networks in people's daily life. The increasing amount of timely microblog data brings new opportunities for enterprises to predict short-term product sales based on microblogs because the daily microblogs posted by various users can express people's sentiments on specific products, such as movies and books. Additionally, the social influence of microblogging platforms enables the rapid spread of product information, implemented by users' forwarding and commenting behavior. To verify the usefulness of microblogs in enhancing the prediction of short-term product sales, in this paper, we first present a new framework that adopts the sentiment and influence features of microblogs. Then, we describe the detailed feature computation methods for sentiment polarity detection and influence measurement. We also implement the Linear Regression (LR) model and the Support Vector Regression (SVR) model, selected as the representatives of linear and nonlinear regression models, to predict short-term product sales. Finally, we take movie box office predictions as an example and conduct experiments to evaluate the performance of the proposed features and models. The results show that the proposed sentiment feature and influence feature of microblogs play a positive role in improving the prediction precision. In addition, both the LR model and the SVR model can lower the MAPE metric of the prediction effectively. [ABSTRACT FROM AUTHOR]

A large body of research uses data from social media websites to predict offline economic outcomes such as sales. However, recent research also points out that such data may be subject to various limitations and biases that may hurt predictive accuracy. At the same time, a growing body of research shows that a new source of online information, search engine logs, has the potential to predict offline outcomes. We study the relationship between these two important data sources in the context of sales predictions. Focusing on the automotive industry, a classic example of a domain of high-involvement products, we use Google’s comprehensive index of Internet discussion forums, in addition to Google search trend data. We find that adding search trend data to models based on the more commonly used social media data significantly improves predictive accuracy. We also find that predictive models based on inexpensive search trend data provide predictive accuracy that is comparable, at least, to that of social media data-based predictive models. Last, we show that the improvement in accuracy is considerably larger for “value” car brands, while for “premium” car brands the improvement obtained is more moderate. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Zhengzhou University: Engineering Science is the property of Editorial Office of Journal of Zhengzhou University: Engineering Science and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Fast food business is the fastest growing business recently in Pakistan and the people need to know the sales forecasting because of the business is totally based on the sales factor sales will tell us how to invest and how to expand the expanse and how to reduce the expanses and the growth of the business is measure in term of sales. We get collected the data form the Lahori fast food chain business and use the base model and find the error MSE rate and then we use the machine leaning mode which is ARIMA to find the MSE error rate and compare their results and finally we found that the base model having the error lower than ARIMA.

This paper presents a study focused on the analysis of phytosanitary treatment sales in the Souss Massa region of Morocco. The objective of the study is to predict the sales of agricultural products, particularly crop protection solutions, aiming to optimize supply chain operations and meet customer demand effectively. Data for this study are collected from multiple sources, including the Enterprise Resource Planning (ERP) system called Microsoft Dynamics AXAPTA used by a leading agricultural company operating in the region. Information such as the date of sale, farming type, climate, and specific sales locations within the Sous Massa region is gathered. Machine learning techniques are applied for forecasting. Various regression models, including the Gradient Boosting Regressor algorithm, are employed to determine the most accurate predictor. Evaluation of the models reveals promising results, with a Mean Absolute Error (MAE) of 0.0035 and a Root Mean Square Error (RMSE) of 0.0066. The results obtained by applying various regression models, including the Gradient Boosting Regressor algorithm, demonstrate promising prediction scores. These findings contribute to the field of sales prediction in the agricultural industry while considering the impact of climate conditions, farming practices, and regional factors. [ABSTRACT FROM AUTHOR]

This study aims to find a robust method to improve the accuracy of online sales prediction. Based on the groundings of existing literature, the authors proposed a Dependency SCOR-topic Sentiment (DSTS) model to analyze the online textual reviews and predict sales performance. The authors took the online sales data of tea as empirical evidence to test the proposed model by integrating the auto-regressive review information model into the DSTS model. The findings include: 1) the effect of the distribution of SCOR-topic from reviews on sales prediction; 2) the effect of review text sentiment on sales prediction increases as the specific topic probability dominates; and 3) the effect of review text sentiment on sales prediction increases as the rest topic probability evenly distributes. These findings demonstrate that the DSTS model is more precise than alternative methods in online sales prediction. This study not only contributes to the literature by pointing out how the distribution of sentiment topic impacts on sales prediction but also has practical implications for the e-commerce practitioners to manage the inventory better and advertise by this prediction method.

Microblogs are one of the major social networks in people’s daily life. The increasing amount of timely microblog data brings new opportunities for enterprises to predict short-term product sales based on microblogs because the daily microblogs posted by various users can express people’s sentiments on specific products, such as movies and books. Additionally, the social influence of microblogging platforms enables the rapid spread of product information, implemented by users’ forwarding and commenting behavior. To verify the usefulness of microblogs in enhancing the prediction of short-term product sales, in this paper, we first present a new framework that adopts the sentiment and influence features of microblogs. Then, we describe the detailed feature computation methods for sentiment polarity detection and influence measurement. We also implement the Linear Regression (LR) model and the Support Vector Regression (SVR) model, selected as the representatives of linear and nonlinear regression models, to predict short-term product sales. Finally, we take movie box office predictions as an example and conduct experiments to evaluate the performance of the proposed features and models. The results show that the proposed sentiment feature and influence feature of microblogs play a positive role in improving the prediction precision. In addition, both the LR model and the SVR model can lower the MAPE metric of the prediction effectively.

Through the rapid developments in machine learning, there have been diverse utilization approaches not only in industrial fields but also in daily life. Implementations of machine learning on financial data, also have been of interest. Herein, we employ machine learning algorithms to store sales data and present future applications for fintech enterprises. We utilize diverse missing data processing methods to handle missing data and apply gradient boosting machine learning algorithms; XGBoost, LightGBM, CatBoost to predict the future revenue of individual stores. As a result, we found that using median imputation onto missing data with the appliance of the xgboost algorithm has the best accuracy. By employing the proposed method, fintech enterprises and customers can attain benefits. Stores can benefit by receiving financial assistance beforehand from fintech companies, while these corporations can benefit by offering financial support to these stores with low risk. [ABSTRACT FROM AUTHOR]

Within the most productive route, online travel agencies (OTAs) intend to use advanced digital media ads to expand their piece of the industry as a whole. The metasearch engine platforms are among the most consistently used digital media environments by OTAs. Most OTAs offer day by day deals in metasearch engine platforms that are paying per click for each hotel to get reservations. The administration of offering methodologies is critical along these lines to reduce costs and increase revenue for online travel agencies. In this study, we tried to predict both the number of impressions and the regular Click-Through-Rate (CTR) level of hotel advertising for each hotel and the daily sales amount. A significant commitment of our research is to use an extended dataset generated by integrating the most informative features implemented in various related studies as the rolling average for a different amount of day and shifted values for use in the proposed test stage for CTR, impression and sales prediction. The data is created in this study by one of Turkey's largest OTA, and we are giving OTA's a genuine application. The results at each prediction stage show that enriching the training data with the OTA-specific additional features, which are the most insightful and sliding window techniques, improves the prediction models ' generalization capability, and tree-based boosting algorithms carry out the greatest results on this problem. Clustering the dataset according to its specifications also improves the results of the predictions. [ABSTRACT FROM AUTHOR]

Background. The explosion of big data (BD), automation, and machine learning have allowed contemporary businesses to better understand and predict human behavior. In scientific research big data have been widely used to study consumer journey and opinions. One of the tools enabling forecasting of sales volume is the Bass diffusion model, which universal nature has been proven in many applications in forecasting the sale of products belonging to various market segments. This article considers the use of BD as exogenous variables in the Bass model to predict the sales of tourist packages. Research aims. The purpose of the research is to assess the impact of using big data on improving the accuracy of forecasts for the sale of tourist packages. The Generalized Bass Model (GBM) has been thus expanded to include big data, which means that exogenous variables include: (1) marketer-generated content (MGC) and (2) user-generated content (UGC), including volume of web search and blog posts. Methodology. This article analyzes online news, blog posts and web search traffic volume related to tourist packages, and then integrates the information into the Bass model, treating it as part of the exogenous variables representing the marketing efforts of tour operators. It has been assumed that the volume of tour operators' web news is a proxy for content generated by marketers (MGC), while the volume of blog posts and web search traffic constitute user-generated content (UGC). Key findings. The empirical analysis found that by incorporating big data into the Bass model provides more accurate prediction of tourist packages' sales volume. In addition, UGC (as an exogenous variable) is better at predicting sales volume than MGC. UGC is a fairly good tool explaining the level of interest and involvement of potential tourists. However, it has been shown that forecasting efficiency is different for blog posts and web search traffic volumes. [ABSTRACT FROM AUTHOR]

The ability to predict product sales is invaluable for improving enterprises' routine decisions of inventory optimization. The most effective solution of sales prediction in real applications is ensemble learning. One challenge of the solution is that it is hard to make timely and accurate predictions because of inadequate decision information and complicated and changeable sales environments. Besides, seeking optimal model combinations from the candidate set is often inaccurate and time-consuming. Another important challenge is that the predicted sales seldom consider "replenishment" of the inventory, which may lead to even higher cost. To address the challenges, we propose a novel dynamic sales prediction model with Auto-Learning and Elastic-Adjustment mechanisms (DSP-FAE): Dynamic sales prediction model can capture dynamic changing patterns of sales time series more effectively. Auto-Learning is used to automatically customize the optimal ensemble learning strategy for each warehouse-product combination in a more efficient way. Elastic-Adjustment is proposed to design a deep neural network-based adjustment factor to correct the predicted sales, which can significantly reduce inventory costs. Extensive offline and online experiments are conducted to verify the performance of the proposed model on two real-world datasets: Galanz and Cainiao. Experimental results show that the proposed DSP-FAE performs better than the selected 10 state-of-the-art baselines significantly in terms of MAE, RRSE and CORR. More importantly, it can save more than 20% inventory cost compared with traditional solutions. [ABSTRACT FROM AUTHOR]