Catalyzing Supply Chain Evolution: A Comprehensive Examination of Artificial Intelligence Integration in Supply Chain Management †
1
Department of Computer Science, Metropolitan College, Boston University, Boston, MA 02215, USA
2
Department of Administration Sciences, Metropolitan College, Boston University, Boston, MA 02215, USA
*
Author to whom correspondence should be addressed.
†
Presented at the 10th International Conference on Time Series and Forecasting, Gran Canaria, Spain, 15–17 July 2024.
Eng. Proc. 2024, 68(1), 57; https://doi.org/10.3390/engproc2024068057
Published: 22 July 2024
(This article belongs to the Proceedings of The 10th International Conference on Time Series and Forecasting)
Download keyboard_arrow_down
Figure 1
<p>Random Under Sampling with Recursive Feature Selection.</p> "> Figure 2
<p>SMOTE with Recursive Feature Selection.</p> "> Figure 3
<p>Data Imbalance Issue in Supply Chain Dataset.</p> "> Figure 4
<p>AUC curve without any oversampling methods.</p> "> Figure 5
<p>Line Order Quantity Forecasting using ARIMA.</p> "> Figure 6
<p>Line Order Quantity Forecasting using SARIMA.</p> ">
Versions Notes
<p>Random Under Sampling with Recursive Feature Selection.</p> "> Figure 2
<p>SMOTE with Recursive Feature Selection.</p> "> Figure 3
<p>Data Imbalance Issue in Supply Chain Dataset.</p> "> Figure 4
<p>AUC curve without any oversampling methods.</p> "> Figure 5
<p>Line Order Quantity Forecasting using ARIMA.</p> "> Figure 6
<p>Line Order Quantity Forecasting using SARIMA.</p> ">
Abstract
:The integration of Artificial Intelligence (AI) into Supply-Chain Management (SCM) has revolutionized operations, offering avenues for enhanced efficiency and decision-making. AI has become pivotal in tackling various Supply-Chain Management challenges, notably enhancing demand forecasting precision and automating warehouse operations for improved efficiency and error reduction. However, a critical debate arises concerning the choice between less accurate explainable models and more accurate yet unexplainable models in Supply-Chain Management applications. This paper explores this debate within the context of various Supply-Chain Management challenges and proposes a methodology for developing models tailored to different Supply-Chain Management problems. Drawing from academic research and modelling, the paper discusses the applications of AI in demand forecasting, inventory optimization, warehouse automation, transportation management, supply chain planning, supplier management, quality control, risk management, and customer service. Additionally, it examines the trade-offs between model interpretability and accuracy, highlighting the need for a nuanced approach. The proposed methodology advocates for the development of explainable models for tasks where interpretability is crucial, such as risk management and supplier selection, while leveraging unexplainable models for tasks prioritizing accuracy, like demand forecasting and predictive maintenance. Through this approach, stakeholders gain insights into Supply-Chain Management processes, fostering better decision-making and accountability.
1. Introduction
The integration of artificial intelligence (AI) technologies into supply chain management (SCM) represents a pivotal advancement in the field, promising enhanced efficiency and decision-making capabilities. This literature survey delves into the applications of AI in supply chain management, synthesizing insights from academic research, scholarly publications, and empirical modeling of supply chain management problem data. Among these applications, demand forecasting and inventory optimization stand out as critical areas undergoing significant transformation. Pioneering studies demonstrate the strides made in AI-driven demand forecasting, resulting in improvements in accuracy and error reduction [1]. Likewise, AI plays a crucial role in optimizing inventory levels, leading to reduced stockouts and enhanced operational efficiency [2,3]. These advancements underscore AI’s practical implications in augmenting demand forecasting and inventory optimization, with specific exploration intended in our study towards order status prediction and time series forecasting. Moreover, AI interventions bolster strategic domains such as supply chain planning and supplier management. By harnessing data on production capacity and supplier capabilities, AI optimizes supply chain strategies and streamlines supplier management processes through automation and data-driven insights, fostering resilience and bolstering supply chain relationships [3,4]. In manufacturing, AI-driven quality control and predictive maintenance solutions herald operational efficiency. AI-powered vision systems improve product quality by detecting defects with precision [5], while predictive maintenance models preemptively identify maintenance needs, minimizing downtime and ensuring consistent product quality [3]. Furthermore, AI’s role in risk management within supply chain operations cannot be overstated. Leveraging advanced analytics, AI enables proactive identification and mitigation of risks, ensuring operational continuity and resilience [6]. Our study aims to augment this understanding by integrating the complex domain of back-order management, enhancing customer satisfaction, and bolstering supply chain resilience.
In Section 2 we discuss the trade-off between accuracy and explainability when it comes to the use of AI in the supply chain. In Section 3 we enumerate the plethora of research available on the utility of AI and machine learning in supply chain management. In Section 4 we posit three specific supply chain issues that can be mitigated using the tools available in machine learning. In Section 5 we reiterate the watershed ideas that form the core of our research.
2. Accuracy vs. Explainability in AI for Supply Chain Management
The ongoing debate regarding the choice between less accurate but explainable models and more accurate yet unexplainable models in AI applications for supply chain management (SCM) is a subject of significant discourse within the field. Explainable models offer transparency and interpretability, allowing stakeholders to comprehend the reasoning behind decisions [7]. However, they may sacrifice accuracy compared to their more complex counterparts. Conversely, unexplainable models, such as deep learning algorithms, often provide superior predictive performance but lack transparency, making it challenging to understand the rationale behind their outputs [8].
Amidst ongoing discourse surrounding the trade-off between accuracy and explainability in AI-driven supply chain management solutions, further research is imperative to develop methodologies that reconcile accuracy and interpretability, ensuring transparency and accountability [7,8]. This issue of balancing accuracy and explainability is particularly crucial in sectors such as healthcare and supply chain management, where the consequences of decision-making can have significant impacts on human lives and operational efficiency. In healthcare, for instance, interpretable models can provide insights into the reasoning behind medical diagnoses and treatment recommendations, fostering trust between healthcare providers and patients. Similarly, in supply chain management, the ability to explain model predictions enables stakeholders to understand the factors driving inventory management decisions, facilitating more informed and effective decision-making processes. Furthermore, the use of feature selection and model weights can enhance explainability by elucidating how specific variables influence model predictions, thereby bridging the gap between accuracy and interpretability.
3. Literature Survey
Section 3 explores the diverse applications of artificial intelligence (AI) in supply chain management (SCM). It begins by highlighting Genetic Algorithms’ (GAs) effectiveness in multi-objective optimization and their role in tackling complex challenges. Natural Language Processing (NLP) streamlines various supply chain management aspects, while Time Series Analysis (TS) optimizes resource allocation and responsiveness. AI and Big Data Analytics (BDA) synergize to enhance resilience, addressing behavioral aspects in decision-making. The section introduces AI’s direct impact on supply chain management performance and resilience, AI Integration (AII) framework dynamics, and commonly used AI techniques.
3.1. Genetic Algorithms in Supply Chain Optimization
Artificial intelligence (AI) has revolutionized supply chain management (SCM) by introducing various techniques, and genetic algorithms (GAs) stand out as powerful tools for multi-objective optimization in supply chain networks. GAs prove effective in partner selection for green supply chain issues, managing multi-product supply chains, and optimizing closed-loop supply chains. The adaptability and efficiency of GAs in addressing complex optimization challenges within supply chain management underscores their potential to revolutionize decision-making processes and enhance overall supply chain efficiency [9].
3.2. Natural Language Processing (NLP) in Supply Chain Management
Natural language processing (NLP) has emerged as a valuable AI technique with diverse applications in supply chain management. From chatbots in marketing campaigns to brand management, customer relationship management, and data collection, NLP automates and streamlines various aspects of supply chain management. It contributes to improved customer interactions, data-driven decision-making, and enhanced operational agility, showcasing its significance within the supply chain management domain [9].
3.3. Time Series Forecasting for Supply-Chain Management Challenges
Time Series Forecasting (TS) is a crucial AI technique addressing a wide spectrum of supply chain management challenges. Its applications span workforce planning, machine scheduling, transport network design, flexible manufacturing, just-in-time production, inventory planning, project portfolio optimization, and route optimization. Harnessing Time series analysis empowers supply chain management practitioners with valuable insights, optimizing resource allocation, and enhancing the responsiveness and adaptability of supply chain operations [9]. In Section 4, the third case study explores the challenges of time series forecasting in supply chain management. By utilizing machine learning techniques such as ARIMA and SARIMA models, the study aims to predict future sales and optimize inventory levels based on historical data. This approach empowers organizations to make informed decisions, reduce uncertainties, and align operations with market demands.
3.4. AI and Big Data Analytics for Supply Chain Resilience
The combined use of artificial intelligence (AI) and big data analytics (BDA) significantly enhances supply chain resilience. AI and big data analytics contribute to the development of dynamic capabilities within firms, facilitating resilience in the face of disruptions. They enhance decision-making processes, particularly during disruptions, by providing insights that clarify uncertainties and reduce risks. Addressing the behavioral aspects of decision-making is crucial when employing AI and big data analytics for supply chain resilience [10].
3.5. AI’s Impact on Supply Chain Performance
AI has a substantial impact on supply chain resilience and performance in dynamic environments. Organizational Information Processing Theory (OIPT), a research framework that highlights the direct impact of AI on supply chain performance in the short term, recommends leveraging AI’s information-processing capabilities to build long-lasting supply chain resilience. The study introduces constructs such as AI capabilities, Adaptive Capabilities (AC), Supply Chain Collaboration (SCC), and Supply Chain Dynamism (SCD) to explain the relationships between AI, performance, and resilience [11].
3.6. AI Integration Framework in Supply Chain Management
The AI Integration (AII) framework provides a theoretical lens for understanding the dynamics of AI integration in supply chain management (SCM). Considering the level of AI integration across the supply chain and its role in decision-making, the AII framework incorporates human sense-making as a factor shaping AI integration and disruption dynamics. It offers insights into the complexities and potential disruptions associated with AI adoption in supply chain management [12].
3.7. Commonly Used AI Techniques in Supply Chain Management
Various artificial intelligence techniques, including Artificial Neural Networks (ANNs), Fuzzy Logic, and Agent-based Systems (ABS)/Multi-Agent Systems (MAS), are prevalent in supply chain management. Machine learning, genetic algorithms, and expert systems find applications in supplier evaluation, supply chain collaboration strategies, and vendor selection. The diverse applications showcase how AI addresses complex decision-making processes in supply chain management [13].
3.8. AI and ML for Manufacturing Enterprise Enhancement
AI and machine learning (ML) play a crucial role in helping manufacturing enterprises enhance profit margins and reduce supply chain costs. These technologies enable the creation of industrial intelligence-driven smart factories, supporting advanced industrial practices through supply chain digital transformation. Predictive analytics, demand forecasting, and real-time decision-making contribute to more efficient and cost-effective supply chain operations [14].
3.9. AI and Blockchain for Supply Chain Resilience
To mitigate supply chain disruptions, the study proposes the use of AI and blockchain technologies. AI-driven data analytics aids in estimating client demand, managing warehouses effectively, and predicting supply chain problems. Blockchain technology enhances supply chain management by minimizing the need for verification, improving functional coordination, inter-firm collaboration, and overall service quality [15].
4. Integrating Machine Learning in Supply Chain Management
Section 4 delves into case studies addressing critical challenges in supply chain management, showcasing the pivotal role of machine learning techniques. Supply chain management and machine learning share a common goal of optimizing processes through data-driven insights. Leveraging machine learning in SCM enhances decision-making, forecasting accuracy, and operational efficiency. By integrating ML algorithms for demand forecasting, inventory optimization, supplier management, and logistics, companies can drive innovation and gain a competitive edge in today’s dynamic business landscape.
4.1. Case Study I: Back-Order Problem
When a customer places an order for a product that is currently unavailable or out of stock, it results in what is known as a back-order [16]. Back-orders can arise due to various reasons, including seasonal fluctuations, unpredictable demand, or logistical issues. While back-ordering can serve to maintain customer satisfaction, ensure a responsive supply chain, and strengthen risk management for companies, inadequate management can lead to tangible and intangible costs such as decreased customer satisfaction, loss of customers, and increased procurement, manufacturing, and delivery costs [17]. In recent times, artificial intelligence (AI) and machine learning (ML) have emerged as prominent tools for forecasting products likely to go out of stock. Deep auto-encoders have been used ubiquitously for back-order prediction and machine learning algorithms are often leveraged to optimize inventory management by determining reordering points [18,19]. Features play a significant role in predicting back-orders by employing various machine learning models such as LightGBM, XGBoost, Logistic Regression, Support Vector Machines, and K-Nearest Neighbors.
Role of Machine Learning
In our study, we explore a diverse range of machine learning algorithms, including Logistic Regression, Decision Tree, Random Forest, and XGBoost [20] using the AUC curve 191 where the diagonal line represents the classifier that makes random guesses where the 192 true positive rate equals the true negative rate (See Figure 1). Each algorithm offers unique advantages and is suitable for different types of data and problem domains. By evaluating and comparing the performance of these algorithms, we gain insights into their effectiveness in predicting backorders and identify the most suitable models for our specific application. Through this comprehensive methodology encompassing sampling, feature selection, and model evaluation using multiple machine learning algorithms, our study aims to develop accurate and robust predictive models for inventory backorders in supply chain management. By leveraging these advanced techniques, organizations can enhance their decision-making processes, optimize inventory management, and ultimately improve customer satisfaction and operational efficiency.
Since accuracy might be misleadingly high when a model only predicts the majority class well and fails to reliably identify the minority class instances, it should not be regarded as a reliable metric in imbalanced class problems. When we examine the results of the above metrics, we generally observe that the model using oversampling, SMOTE, provided higher F1 and AUC scores (see Figure 2). It is seen that in both resampling methods, more complex algorithms such as Random Forest and Gradient Boosting are the most successful algorithms. This situation can also be said to result in a trade-off between the explainability of simpler algorithms and the higher accuracy of more complex algorithms. Therefore, the choice of algorithm to be used is related to which of these advantages the company wants to benefit from.
4.2. Case Study II: Order Status Prediction
The modern supply chain is a complex and interconnected network that spans the globe, involving numerous stakeholders and intricate processes. One of the critical challenges faced by supply chain managers is the accurate prediction of order status. The ability to forecast the status of orders in advance has a profound impact on operational efficiency, customer satisfaction, and overall business performance. Machine learning, with its predictive analytics capabilities, has emerged as a transformative tool in addressing this challenge.
The order status prediction problem revolves around anticipating the state of an order at different stages of its lifecycle within the supply chain. These stages include order placement, processing, packaging, shipping, and delivery. Predicting the order status with precision allows supply chain managers to proactively identify potential delays, optimize resource allocation, and enhance the overall responsiveness of the supply chain. Several factors contribute to the complexity of order status prediction in the supply chain. Variability in demand, unexpected disruptions, and the dynamic nature of supply chain processes pose significant challenges. Traditional methods often struggle to adapt to these complexities, making machine learning an attractive solution.
Role of Machine Learning
To illustrate the utility of machine learning we leverage a dataset with relevant features such as days for shipping, benefit per order, and product price; the classifiers are trained on historical data to forecast the delivery status of orders. Before directly using the data as input to our prediction models, we evaluated the count of each class in the delivery status to take into account the class imbalance problem (see Figure 3). The count of classes indicates that there exists an imbalance with the classes. To address the issue, we have used multiple techniques such as the Synthetic Minority Over-sampling Technique (SMOTE), BorderlineSMOTE, and ADYSN and class weight adjustment for oversampling to mitigate the potential biases that might surface in the results provided by the machine learning models [21,22]. The classifiers employed include Decision Trees, Random Forests, Logistic Regression, K-Nearest Neighbors, Gradient Boosting, and Naive Bayes. The Multi-Class ROC-AUC Curves visually showcase the performance of each classifier. The Area Under the Curve (AUC) metrics quantitatively evaluate the classifiers’ ability to distinguish between different delivery status classes. The higher the AUC value, the better the model’s predictive accuracy (see Figure 4).
The comprehensive analysis of Multi-Class AUC values across various classification models and oversampling techniques provides valuable insights into the impact of class imbalance mitigation on model performance. Notably, Random Forest, Gradient Boosting, MLP, and AdaBoost consistently exhibit high AUC values across all scenarios, showcasing their robustness in handling imbalanced datasets and maintaining discriminating power. Logistic Regression and Linear Discriminant Analysis (LDA) demonstrate stability and resilience to class imbalance, while K-Nearest Neighbors (KNN) appears sensitive to imbalanced class distributions, with a noticeable decrease in AUC values. The effectiveness of oversampling techniques such as SMOTE and ADASYN, coupled with class weight adjustment, varies across models, emphasizing the importance of algorithm choice in imbalanced settings. Naive Bayes consistently performs well but shows a slight decrease in AUC values with oversampling techniques. This nuanced understanding of model behavior guides practitioners in selecting appropriate strategies for improved multi-class classification outcomes.
4.3. Case Study III: Time Series Forecasting
In the realm of supply chain management, the time forecasting problem is a critical challenge that revolves around predicting future sales or order quantities based on historical data. Leveraging machine learning techniques, this endeavor seeks to harness the power of algorithms and predictive models to analyze past patterns and trends. By delving into historical data, these models aim to uncover hidden insights that can inform accurate predictions regarding future demand. The time forecasting problem is particularly crucial for supply chain managers who strive to optimize inventory levels, streamline production processes, and enhance overall operational efficiency. Through the application of machine learning, organizations can move beyond traditional forecasting methods, incorporating advanced algorithms that adapt to changing market dynamics and unforeseen disruptions. Ultimately, addressing the time forecasting problem with machine learning in supply chain management empowers businesses to make informed decisions, reduce uncertainties, and align their operations with the ever-evolving demands of the market.
Role of Machine Learning
Machine learning addresses the time forecasting problem in supply chain management using two distinct approaches: ARIMA and SARIMA models. ARIMA is a powerful model designed for time series analysis, incorporating three main components: AutoRegressive (AR), Integrated (I), and Moving Average (MA). The AR component captures the relationship between an observation and its previous values, the I component involves applying the difference operation on the series to achieve a stationary state, and the MA component models the dependency between an observation and a residual error from the moving average. SARIMA extends the capabilities of ARIMA by incorporating seasonality, making it particularly suitable for datasets exhibiting recurring patterns at fixed intervals. The ’S’ in SARIMA stands for ’Seasonal,’ and it introduces additional parameters to account for periodic fluctuations in the data. The inclusion of seasonal components allows SARIMA to address time series with a clear seasonal structure, such as monthly or quarterly patterns [23]. Distinguishing between ARIMA and SARIMA, the latter is more adept at handling datasets with identifiable seasonal trends. ARIMA, on the other hand, is suitable for non-seasonal time series data. Both models are widely used in supply chain management for their ability to uncover underlying patterns and trends within historical data, facilitating accurate predictions of future demand [24]. We applied the ARIMA and SARIMA models to the order quantity data (see Figure 5 and Figure 6). The model is configured with the relevant order parameter to predict the value of line order quantity for the next eight years from the historical data.
5. Conclusions
In conclusion, the integration of artificial intelligence in supply chain management presents a transformative opportunity to enhance operational efficiency, decision-making capabilities, and overall supply chain resilience. By leveraging AI technologies for demand forecasting, inventory optimization, supplier management, and risk mitigation, organizations can streamline processes, reduce costs, and improve customer satisfaction. However, it is crucial to carefully balance the trade-off between accuracy and explainability in AI models to ensure transparency and accountability in decision-making processes. Moving forward, continued research and innovation in AI integration hold the key to catalyzing supply chain evolution and driving sustainable growth in the industry.
Author Contributions
All authors contributed equally to this paper. All authors have read and agreed to the published version of the manuscript.
Funding
The research received no external funding.
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
The sentencing data is located at https://www.kaggle.com/datasets/shashwatwork/dataco-smart-supply-chain-for-big-data-analysis, accessed on 25 January 2023.
Acknowledgments
The authors would like to thank the Metropolitan College of Boston University for their support.
Conflicts of Interest
The authors declare no conflict of interest.
References
- Kazemi, M.; Rahimi, N. A Comprehensive Review on Forecasting Demand in Supply Chain Management. Comput. Ind. Eng. 2019, 135, 1272–1285. [Google Scholar]
- Feng, Q.; Li, X.; Liu, J.; Wang, Z.; Wang, Z. A Survey of Demand Forecasting in Fashion Retail. IEEE Trans. Ind. Inform. 2020, 17, 4471–4481. [Google Scholar]
- Li, S.; Fu, Y.; Wang, J.; Zhang, S. Optimizing the Green Supply Chain Coordination for Perishable Products under Asymmetric Information and Supply Disruption. Eur. J. Oper. Res. 2020, 283, 625–642. [Google Scholar]
- Zhang, Y.; Liu, Y.; Wan, L. Supplier Selection in Fashion Retailing: A Data-Driven Approach. IEEE Trans. Ind. Inform. 2019, 16, 6549–6558. [Google Scholar]
- Gupta, R.; Dahiya, A.; Kumar, V.; Jain, A.; Singh, P.K. Artificial Intelligence in Vision Based Quality Control: A Review. Robot. Comput.-Integr. Manuf. 2021, 69, 102228. [Google Scholar]
- Chen, J.; Shen, Z.J.M.; Wang, J. An Ensemble Learning Method for Supply Chain Disruption Prediction. Eur. J. Oper. Res. 2020, 283, 1077–1091. [Google Scholar]
- Ribeiro, M.T.; Singh, S.; Guestrin, C. “Why should I trust you?”: Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, 13–17 August 2016; pp. 1135–1144. [Google Scholar]
- Lipton, Z.C. The Mythos of Model Interpretability. arXiv 2016, arXiv:1606.03490. [Google Scholar]
- Toorajipour, R.; Sohrabpour, V.; Nazarpour, A.; Oghazi, P.; Fischl, M. Artificial intelligence in supply chain management: A systematic literature review. J. Bus. Res. 2021, 122, 502–517. [Google Scholar] [CrossRef]
- Zamani, E.; Smyth, C.; Gupta, S.; Dennehy, D. Artificial Intelligence and Big Data Analytics for Supply Chain Resilience: A Systematic Literature Review. Ann. Oper. Res. 2022, 327, 605–632. [Google Scholar] [CrossRef] [PubMed]
- Belhadi, A.; Mani, V.; Kamble, S.; Khan, S.; Verma, S. Artificial intelligence-driven innovation for enhancing supply chain resilience and performance under the effect of supply chain dynamism: An empirical investigation. Ann. Oper. Res. 2021, 333, 627–652. [Google Scholar] [CrossRef] [PubMed]
- Hendriksen, C. Artificial intelligence for supply chain management: Disruptive innovation or innovative disruption? J. Supply Chain. Manag. 2023, 59, 65–76. [Google Scholar] [CrossRef]
- Atwani, M.; Hlyal, M.; Elalami, J. A Review of Artificial Intelligence applications in Supply Chain. ITM Web Conf. 2022, 46, 03001. [Google Scholar] [CrossRef]
- Rana, J.; Daultani, Y. Mapping the Role and Impact of Artificial Intelligence and Machine Learning Applications in Supply Chain Digital Transformation: A Bibliometric Analysis. Oper. Manag. Res. 2023, 16, 1641–1666. [Google Scholar] [CrossRef]
- Kashem, M.A.; Shamsuddoha, M.; Nasir, T.; Chowdhury, A.A. Supply Chain Disruption versus Optimization: A Review on Artificial Intelligence and Blockchain. Knowledge 2023, 3, 80–96. [Google Scholar] [CrossRef]
- Shajalal, M.; Hajek, P.; Abedin, M. Product Backorder Prediction Using Deep Neural Network on Imbalanced Data. Int. J. Prod. Res. 2021, 61, 302–319. [Google Scholar] [CrossRef]
- Hajek, P.; Abedin, M.Z. A Profit Function-Maximizing Inventory Backorder Prediction System using Big Data Analytics. IEEE Access 2020, 8, 58982–58994. [Google Scholar] [CrossRef]
- Saraogi, G.; Gupta, D.; Sharma, L.; Rana, A. An Un-Supervised Approach for Backorder Prediction Using Deep Autoencoder. Recent Adv. Comput. Sci. Commun. 2021, 14, 500–511. [Google Scholar] [CrossRef]
- Inprasit, T.; Tanachutiwat, S. Reordering Point Determination Using Machine Learning Technique for Inventory Management. In Proceedings of the International Conference on Engineering, Applied Sciences, and Technology (ICEAST), Phuket, Thailand, 4–7 July 2018; pp. 1–4. [Google Scholar]
- Chen, T.; Guestrin, C. XGBoost: A Scalable Tree Boosting System. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD’16, San Francisco, CA, USA, 13–17 August 2016; Association for Computing Machinery: New York, NY, USA, 2016; pp. 785–794. [Google Scholar] [CrossRef]
- Wang, S.; Dai, Y.; Shen, J.; Xuan, J. Research on expansion and classification of imbalanced data based on SMOTE algorithm. Sci. Rep. 2021, 11, 24039. [Google Scholar] [CrossRef] [PubMed]
- Chawla, N.; Bowyer, K.; Hall, L.; Kegelmeyer, W. SMOTE: Synthetic Minority Over-sampling Technique. J. Artif. Intell. Res. (JAIR) 2002, 16, 321–357. [Google Scholar] [CrossRef]
- Permanasari, A.; Hidayah, I.; Bustoni, I.A. SARIMA (Seasonal ARIMA) implementation on time series to forecast the number of Malaria incidence. In Proceedings of the 2013 International Conference on Information Technology and Electrical Engineering (ICITEE), Yogyakarta, Indonesia, 7–8 October 2013; pp. 203–207. [Google Scholar] [CrossRef]
- Chhabra, H. A Comparative Study of ARIMA and SARIMA Models to Forecast Lockdowns Due to COVID-19 (version 1). 2022; Preprint. [Google Scholar] [CrossRef]
Figure 1.
Random Under Sampling with Recursive Feature Selection.
Figure 2.
SMOTE with Recursive Feature Selection.
Figure 3.
Data Imbalance Issue in Supply Chain Dataset.
Figure 4.
AUC curve without any oversampling methods.
Figure 5.
Line Order Quantity Forecasting using ARIMA.
Figure 6.
Line Order Quantity Forecasting using SARIMA.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
MDPI and ACS Style
Pattnaik, S.; Liew, N.; Kures, A.O.; Pinsky, E.; Park, K. Catalyzing Supply Chain Evolution: A Comprehensive Examination of Artificial Intelligence Integration in Supply Chain Management. Eng. Proc. 2024, 68, 57. https://doi.org/10.3390/engproc2024068057
AMA Style
Pattnaik S, Liew N, Kures AO, Pinsky E, Park K. Catalyzing Supply Chain Evolution: A Comprehensive Examination of Artificial Intelligence Integration in Supply Chain Management. Engineering Proceedings. 2024; 68(1):57. https://doi.org/10.3390/engproc2024068057
Chicago/Turabian StylePattnaik, Sarthak, Natasya Liew, Ali Ozcan Kures, Eugene Pinsky, and Kathleen Park. 2024. "Catalyzing Supply Chain Evolution: A Comprehensive Examination of Artificial Intelligence Integration in Supply Chain Management" Engineering Proceedings 68, no. 1: 57. https://doi.org/10.3390/engproc2024068057