Alibaba has developed a series of models and algorithms that include deep-learning algorithms for demand forecasting, simulation-optimization-based models for inventory management, price optimization for promotions, and markdown optimization for product recommendations. These models and algorithms have been implemented in its omnichannel retail infrastructure of several retail business subsidiaries over the past three years, and have generated more than 100 million dollars in annual cost reductions in cost and increases in profit.
{"title":"Alibaba Realizes Millions in Cost Savings Through Integrated Demand Forecasting, Inventory Management, Price Optimization, and Product Recommendations","authors":"Yuming Deng, Xinhui Zhang, Tong Wang, Lung-Chuang Wang, Yidong Zhang, Xiaoqing Wang, Su Zhao, Yunwei Qi, Guangyao Yang, Xuezheng Peng","doi":"10.1287/inte.2022.1145","DOIUrl":"https://doi.org/10.1287/inte.2022.1145","url":null,"abstract":"Alibaba has developed a series of models and algorithms that include deep-learning algorithms for demand forecasting, simulation-optimization-based models for inventory management, price optimization for promotions, and markdown optimization for product recommendations. These models and algorithms have been implemented in its omnichannel retail infrastructure of several retail business subsidiaries over the past three years, and have generated more than 100 million dollars in annual cost reductions in cost and increases in profit.","PeriodicalId":53206,"journal":{"name":"Informs Journal on Applied Analytics","volume":"32 1","pages":""},"PeriodicalIF":1.4,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75155969","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
D. Bertsimas, Michael Lingzhi Li, Xinggang Liu, Jennings Xu, Najat Khan
The COVID-19 pandemic has spurred extensive vaccine research worldwide. One crucial part of vaccine development is the phase III clinical trial that assesses the vaccine for safety and efficacy in the prevention of COVID-19. In this work, we enumerate the first successful implementation of using machine learning models to accelerate phase III vaccine trials, working with the single-dose Johnson & Johnson vaccine to predictively select trial sites with naturally high incidence rates (“hotspots”). We develop DELPHI, a novel, accurate, policy-driven machine learning model that serves as the basis of our predictions. During the second half of 2020, the DELPHI-driven site selection identified hotspots with more than 90% accuracy, shortened trial duration by six to eight weeks (approximately 33%), and reduced enrollment by 15,000 (approximately 25%). In turn, this accelerated time to market enabled Janssen’s vaccine to receive its emergency use authorization and realize its public health impact earlier than expected. Several geographies identified by DELPHI have since been the first areas to report variants of concern (e.g., Omicron in South Africa), and thus DELPHI’s choice of these areas also produced early data on how the vaccine responds to new threats. Johnson & Johnson has also implemented a similar approach across its business including supporting trial site selection for other vaccine programs, modeling surgical procedure demand for its Medical Device unit, and providing guidance on return-to-work programs for its 130,000 employees. Continued application of this methodology can help shorten clinical development and change the economics of drug development by reducing the level of risk and cost associated with investing in novel therapies. This will allow Johnson & Johnson and others to enable more effective delivery of medicines to patients. Funding: This work was funded by Janssen Research & Development, LLC.
{"title":"Data-Driven COVID-19 Vaccine Development for Janssen","authors":"D. Bertsimas, Michael Lingzhi Li, Xinggang Liu, Jennings Xu, Najat Khan","doi":"10.1287/inte.2022.1150","DOIUrl":"https://doi.org/10.1287/inte.2022.1150","url":null,"abstract":"The COVID-19 pandemic has spurred extensive vaccine research worldwide. One crucial part of vaccine development is the phase III clinical trial that assesses the vaccine for safety and efficacy in the prevention of COVID-19. In this work, we enumerate the first successful implementation of using machine learning models to accelerate phase III vaccine trials, working with the single-dose Johnson & Johnson vaccine to predictively select trial sites with naturally high incidence rates (“hotspots”). We develop DELPHI, a novel, accurate, policy-driven machine learning model that serves as the basis of our predictions. During the second half of 2020, the DELPHI-driven site selection identified hotspots with more than 90% accuracy, shortened trial duration by six to eight weeks (approximately 33%), and reduced enrollment by 15,000 (approximately 25%). In turn, this accelerated time to market enabled Janssen’s vaccine to receive its emergency use authorization and realize its public health impact earlier than expected. Several geographies identified by DELPHI have since been the first areas to report variants of concern (e.g., Omicron in South Africa), and thus DELPHI’s choice of these areas also produced early data on how the vaccine responds to new threats. Johnson & Johnson has also implemented a similar approach across its business including supporting trial site selection for other vaccine programs, modeling surgical procedure demand for its Medical Device unit, and providing guidance on return-to-work programs for its 130,000 employees. Continued application of this methodology can help shorten clinical development and change the economics of drug development by reducing the level of risk and cost associated with investing in novel therapies. This will allow Johnson & Johnson and others to enable more effective delivery of medicines to patients. Funding: This work was funded by Janssen Research & Development, LLC.","PeriodicalId":53206,"journal":{"name":"Informs Journal on Applied Analytics","volume":"17 1","pages":""},"PeriodicalIF":1.4,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90350718","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Peiling Wu-Smith, P. Keenan, Jonathan H. Owen, Andrew Norton, Kelly Kamm, Kathryn M. Schumacher, P. Fenyes, Don Kiggins, Philip W. Konkel, W. Rosen, Kurt Schmitter, Sharon Sheremet, Laura Yochim
General Motors (GM) vehicles have more than 100 customer-facing features, known as vehicle content. Decisions about how to package and price these features have a significant impact on our customers’ experiences and on GM’s business results. Vehicle features are assigned as standard, optional, or unavailable on different trim levels, resulting in an enormous combinatorial solution space. Vehicle content optimization (VCO) combines customer market research, discrete choice models, and custom multiobjective nonlinear optimization algorithms to optimize vehicle contenting and pricing decisions. VCO comprehends complex dynamics and tradeoffs and allows GM to optimally balance customer preferences and profitability. After six years of development and multiple proof-of-concept and pilot studies, VCO was officially integrated into GM’s Global Vehicle Development Process in 2014. As of 2021, VCO has been used on more than 85 vehicle programs globally. It has enabled customer-centric product development and more efficient engineering, sourcing, and manufacturing. GM Finance verified that VCO enabled $4.4 billion of incremental profit over the average product life cycle (i.e., six years on average) since 2018, making it a vastly impactful example of operations research and applied analytics.
{"title":"General Motors Optimizes Vehicle Content for Customer Value and Profitability","authors":"Peiling Wu-Smith, P. Keenan, Jonathan H. Owen, Andrew Norton, Kelly Kamm, Kathryn M. Schumacher, P. Fenyes, Don Kiggins, Philip W. Konkel, W. Rosen, Kurt Schmitter, Sharon Sheremet, Laura Yochim","doi":"10.1287/inte.2022.1144","DOIUrl":"https://doi.org/10.1287/inte.2022.1144","url":null,"abstract":"General Motors (GM) vehicles have more than 100 customer-facing features, known as vehicle content. Decisions about how to package and price these features have a significant impact on our customers’ experiences and on GM’s business results. Vehicle features are assigned as standard, optional, or unavailable on different trim levels, resulting in an enormous combinatorial solution space. Vehicle content optimization (VCO) combines customer market research, discrete choice models, and custom multiobjective nonlinear optimization algorithms to optimize vehicle contenting and pricing decisions. VCO comprehends complex dynamics and tradeoffs and allows GM to optimally balance customer preferences and profitability. After six years of development and multiple proof-of-concept and pilot studies, VCO was officially integrated into GM’s Global Vehicle Development Process in 2014. As of 2021, VCO has been used on more than 85 vehicle programs globally. It has enabled customer-centric product development and more efficient engineering, sourcing, and manufacturing. GM Finance verified that VCO enabled $4.4 billion of incremental profit over the average product life cycle (i.e., six years on average) since 2018, making it a vastly impactful example of operations research and applied analytics.","PeriodicalId":53206,"journal":{"name":"Informs Journal on Applied Analytics","volume":"72 1","pages":""},"PeriodicalIF":1.4,"publicationDate":"2023-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87845072","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Decision analysis is widely used in making decisions involving low-probability, high-consequence events (e.g., drug discovery, oil and gas drilling, risk reduction). This paper focuses on the automotive industries in which more intermediate uncertainties are important. As in any large organization, different members of the organization have different information and different incentives. In this setting, influence diagrams proved invaluable in identifying the information creditable models need, discovering new distinctions of high value, developing win/win compromises, and enabling higher-value technology transfer. However, these examples also highlight the need for more research on addressing motivational biases within organizations. History: This paper was refereed. This paper was accepted for the Special Issue of INFORMS Journal on Applied Analytics—Decision Analysis
{"title":"Lessons for Decision-Analysis Practice from the Automotive Industry","authors":"R. Bordley","doi":"10.1287/inte.2022.1151","DOIUrl":"https://doi.org/10.1287/inte.2022.1151","url":null,"abstract":"Decision analysis is widely used in making decisions involving low-probability, high-consequence events (e.g., drug discovery, oil and gas drilling, risk reduction). This paper focuses on the automotive industries in which more intermediate uncertainties are important. As in any large organization, different members of the organization have different information and different incentives. In this setting, influence diagrams proved invaluable in identifying the information creditable models need, discovering new distinctions of high value, developing win/win compromises, and enabling higher-value technology transfer. However, these examples also highlight the need for more research on addressing motivational biases within organizations. History: This paper was refereed. This paper was accepted for the Special Issue of INFORMS Journal on Applied Analytics—Decision Analysis","PeriodicalId":53206,"journal":{"name":"Informs Journal on Applied Analytics","volume":"42 1","pages":""},"PeriodicalIF":1.4,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91332290","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S. Kulturel-Konak, A. Konak, Lily Jakielaszek, N. Gavirneni
Continuing care facilities are a rapidly growing segment of senior living communities providing end-to-end solutions comprising independent living, assisted living, nursing home care, and ultimately hospice. All these establishments contain (in addition to other facilities associated with living, exercising, learning, activities, etc.) dining services managed by an interdisciplinary (finance, nutrition, dietitian, kitchen operations, hospitality, and procurement) team of executives, each with their own objective while cognizant of the overarching organizational, operational, and financial metrics. The residents of these facilities consume most of their meals at these dining facilities, necessitating that the food served meets the complete nutrition, dietary, cost, and operational requirements. Thus, the menu (often rotating every few weeks) of food items must be carefully chosen to be efficiently procured, processed, and served, all the while meeting the nutritional, dietary, and patron satisfaction constraints each put forth by the corresponding stakeholder. We address this complex, unwieldy, and large multiobjective optimization problem using mixed integer linear programming. We demonstrate how menu planners and chefs can analyze their decisions regarding menu structures and evaluate alternative menu interventions to improve menus’ nutritional value while ensuring their residents’ autonomy in making food choice decisions. Along the way, we interviewed various stakeholders, identified their objectives and constraints, gathered the necessary data, formulated and solved the resulting optimization problems, and produced demonstrably effective menus.
{"title":"Menu Engineering for Continuing Care Senior Living Facilities with Captive Dining Patrons","authors":"S. Kulturel-Konak, A. Konak, Lily Jakielaszek, N. Gavirneni","doi":"10.1287/inte.2022.1140","DOIUrl":"https://doi.org/10.1287/inte.2022.1140","url":null,"abstract":"Continuing care facilities are a rapidly growing segment of senior living communities providing end-to-end solutions comprising independent living, assisted living, nursing home care, and ultimately hospice. All these establishments contain (in addition to other facilities associated with living, exercising, learning, activities, etc.) dining services managed by an interdisciplinary (finance, nutrition, dietitian, kitchen operations, hospitality, and procurement) team of executives, each with their own objective while cognizant of the overarching organizational, operational, and financial metrics. The residents of these facilities consume most of their meals at these dining facilities, necessitating that the food served meets the complete nutrition, dietary, cost, and operational requirements. Thus, the menu (often rotating every few weeks) of food items must be carefully chosen to be efficiently procured, processed, and served, all the while meeting the nutritional, dietary, and patron satisfaction constraints each put forth by the corresponding stakeholder. We address this complex, unwieldy, and large multiobjective optimization problem using mixed integer linear programming. We demonstrate how menu planners and chefs can analyze their decisions regarding menu structures and evaluate alternative menu interventions to improve menus’ nutritional value while ensuring their residents’ autonomy in making food choice decisions. Along the way, we interviewed various stakeholders, identified their objectives and constraints, gathered the necessary data, formulated and solved the resulting optimization problems, and produced demonstrably effective menus.","PeriodicalId":53206,"journal":{"name":"Informs Journal on Applied Analytics","volume":"1 1","pages":""},"PeriodicalIF":1.4,"publicationDate":"2022-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90475323","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As the frequency and duration of grid outages increase, backup power systems are becoming more important for ensuring that critical infrastructure continues to provide essential services. Most facilities rely on diesel generators, which may be ineffective during long outages owing to limited fuel supplies and high generator failure rates. Distributed energy resources such as solar, storage, and combined-heat-and-power systems, coupled with on-site biofuel production, offer an alternative source of on-site generation that can provide both cost savings and resilience (i.e., the ability to respond to catastrophic events with longer-term consequences). A mixed-integer linear program minimizes costs and maximizes resilience at a wastewater treatment plant in Wilmington, North Carolina. We find that the plant can reduce life-cycle energy costs by 3.1% through the installation of a hybrid combined-heat-and-power, photovoltaic, and storage system. When paired with existing diesel generators, this system can sustain full load for seven days while saving $664,000 over 25 years and reducing diesel fuel use by 48% compared with the diesel-only solution. This analysis informed a decision by the Cape Fear Public Utility Authority to allocate funds for the implementation of a combined-heat-and-power system at the wastewater treatment plant in fiscal year 2023. The benefits of deploying hybrid combined-heat-and-power technologies and the utilization of on-site biofuel production extend, on a national scale, to thousands of wastewater treatment facilities and other types of critical infrastructure.
{"title":"North Carolina Water Utility Builds Resilience with Distributed Energy Resources","authors":"Kate Anderson, James Grymes, A. Newman, A. Warren","doi":"10.1287/inte.2022.1136","DOIUrl":"https://doi.org/10.1287/inte.2022.1136","url":null,"abstract":"As the frequency and duration of grid outages increase, backup power systems are becoming more important for ensuring that critical infrastructure continues to provide essential services. Most facilities rely on diesel generators, which may be ineffective during long outages owing to limited fuel supplies and high generator failure rates. Distributed energy resources such as solar, storage, and combined-heat-and-power systems, coupled with on-site biofuel production, offer an alternative source of on-site generation that can provide both cost savings and resilience (i.e., the ability to respond to catastrophic events with longer-term consequences). A mixed-integer linear program minimizes costs and maximizes resilience at a wastewater treatment plant in Wilmington, North Carolina. We find that the plant can reduce life-cycle energy costs by 3.1% through the installation of a hybrid combined-heat-and-power, photovoltaic, and storage system. When paired with existing diesel generators, this system can sustain full load for seven days while saving $664,000 over 25 years and reducing diesel fuel use by 48% compared with the diesel-only solution. This analysis informed a decision by the Cape Fear Public Utility Authority to allocate funds for the implementation of a combined-heat-and-power system at the wastewater treatment plant in fiscal year 2023. The benefits of deploying hybrid combined-heat-and-power technologies and the utilization of on-site biofuel production extend, on a national scale, to thousands of wastewater treatment facilities and other types of critical infrastructure.","PeriodicalId":53206,"journal":{"name":"Informs Journal on Applied Analytics","volume":"78 1","pages":""},"PeriodicalIF":1.4,"publicationDate":"2022-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73165970","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
C. Subbiah, Andrea C. Hupman, Haitao Li, Joseph P. Simonis
A U.S. Midwestern Fortune 500 financial services firm develops software capabilities in-house and requires predictions of project needs for efficient resource allocation decisions across the many projects operating simultaneously. The company develops a novel prediction tool based on the projects’ required software development tasks as described by firm-specific design patterns. The firm provides these predictions within a set of estimates based on industry standard function count methods as well as firm-specific predictive models based on function points and on initial labor assignments. Company management is thus equipped with predictions from multiple methodologies and multiple information sources, enhancing the firm’s ability to predict project needs. Managers aggregate the forecasts, with prediction performance estimated to improve by 35%–49%, measured relative to estimates of the absolute percentage error of the prior method. The improved predictions provide a significant advantage to planning decisions and efficient internal operations. Insights to how managers aggregate the set of forecasts and insights to how the models contribute to the scaled value of information are discussed and further illustrate the benefits of the approach.
{"title":"Improving Software Development Effort Estimation with a Novel Design Pattern Model","authors":"C. Subbiah, Andrea C. Hupman, Haitao Li, Joseph P. Simonis","doi":"10.1287/inte.2022.1138","DOIUrl":"https://doi.org/10.1287/inte.2022.1138","url":null,"abstract":"A U.S. Midwestern Fortune 500 financial services firm develops software capabilities in-house and requires predictions of project needs for efficient resource allocation decisions across the many projects operating simultaneously. The company develops a novel prediction tool based on the projects’ required software development tasks as described by firm-specific design patterns. The firm provides these predictions within a set of estimates based on industry standard function count methods as well as firm-specific predictive models based on function points and on initial labor assignments. Company management is thus equipped with predictions from multiple methodologies and multiple information sources, enhancing the firm’s ability to predict project needs. Managers aggregate the forecasts, with prediction performance estimated to improve by 35%–49%, measured relative to estimates of the absolute percentage error of the prior method. The improved predictions provide a significant advantage to planning decisions and efficient internal operations. Insights to how managers aggregate the set of forecasts and insights to how the models contribute to the scaled value of information are discussed and further illustrate the benefits of the approach.","PeriodicalId":53206,"journal":{"name":"Informs Journal on Applied Analytics","volume":"17 1","pages":""},"PeriodicalIF":1.4,"publicationDate":"2022-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76642884","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Implementations of new technology in large organizations are often plagued with delays and dissatisfaction among customers, employees, and management. At Capital One, group decision theory was used to provide an assessment of the risks of three implementation strategies from a company-wide perspective, along with the risks to each of the stakeholder groups within the company. The analysis revealed that each of the three alternatives would negatively affect one or more of the stakeholder groups, leading to the development of a new implementation strategy which allowed for a successful rollout avoiding delays that plagued other banks implementing the same technology and improving both employee and customer satisfaction.
{"title":"Technology Implementation at Capital One","authors":"E. Cook, Jason R. W. Merrick","doi":"10.1287/inte.2022.1135","DOIUrl":"https://doi.org/10.1287/inte.2022.1135","url":null,"abstract":"Implementations of new technology in large organizations are often plagued with delays and dissatisfaction among customers, employees, and management. At Capital One, group decision theory was used to provide an assessment of the risks of three implementation strategies from a company-wide perspective, along with the risks to each of the stakeholder groups within the company. The analysis revealed that each of the three alternatives would negatively affect one or more of the stakeholder groups, leading to the development of a new implementation strategy which allowed for a successful rollout avoiding delays that plagued other banks implementing the same technology and improving both employee and customer satisfaction.","PeriodicalId":53206,"journal":{"name":"Informs Journal on Applied Analytics","volume":"13 1","pages":""},"PeriodicalIF":1.4,"publicationDate":"2022-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81105826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：