Decision Analytics Journal最新文献

{"title":"An optimal imputation algorithm for reducing bias and errors in missing data handling for AI models","authors":"Anu Maria Sebastian ,&nbsp;David Peter ,&nbsp;Rinu Ann Sebastian","doi":"10.1016/j.dajour.2025.100627","DOIUrl":"10.1016/j.dajour.2025.100627","url":null,"abstract":"<div><div>Data is an essential fuel for artificial intelligence (AI) to power the underlying machine learning (ML) algorithms. Missing data is common in most real-world datasets due to measurement errors, non-responses, and human errors during the data collection, which can ultimately lead to reduced accuracy and reliability for the AI models. Moreover, many ML algorithms are designed to work with complete datasets. Data imputation (DI) assists in creating a comprehensive representation of the data, allowing AI models to learn from an affluent dataset and generate more accurate results. Therefore, choosing the proper imputation technique is essential in minimizing the errors and biases introduced in the data during the imputation. The difficulty in creating an imputation method that performs optimally across the entire spectrum of data stems from the disparity in the inherent characteristics displayed by the different datasets. All the existing DI selection approaches are computationally intensive, demanding repetitive and exhaustive experimentation of the popular DI methods on every new dataset to evaluate its suitability, resulting in significant wastage of time and effort. This research proposes an algorithm for systematically selecting an optimal imputation technique based on the intrinsic characteristics of the dataset. It associates the performance of DI algorithms with the specific characteristics of a given dataset using a characteristics chart (C-chart). The resulting DI recommendation will remain valid for another dataset with a similar C-chart. Thus, our method eliminates the need for exhaustive experimentation to find the proper DI method and offers a reliable imputation for real-world datasets that lack a verifiable ground truth. We have demonstrated the performance of our method using a suite of six benchmark DI algorithms, eight public datasets, and two ML classifiers. We use both Normalized Root Mean Square Error (NRMSE) and Jensen Shannon Distance (JSD) scores to evaluate the potential of the DI algorithms. We could observe that the recommended DI algorithms could enhance the ML classifier accuracy by up to 19.8%. We believe that the proposed algorithm is a significant step towards automating the selection of an optimal DI technique based on data characteristics.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100627"},"PeriodicalIF":0.0,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144925152","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A decision support framework using pruned trees for analytical quality assessment in agri-marine products","authors":"Wilma Latuny ,&nbsp;Victor Oryon Lawalata ,&nbsp;Geovanny Branchiny Imasuly","doi":"10.1016/j.dajour.2025.100628","DOIUrl":"10.1016/j.dajour.2025.100628","url":null,"abstract":"<div><div>Seaweed is a globally significant aquaculture commodity with increasing economic and environmental importance. However, inconsistent quality standards and subjective procurement practices continue to hinder efficiency in its value chain, particularly for dried Eucheuma seaweed. This study presents a decision support system that applies the C4.5 decision tree algorithm, enhanced through six pruning techniques, to classify seaweed quality and guide purchasing decisions in informal market settings. A dataset of 259 entries was compiled, capturing nine key quality attributes. This dataset was used to develop and evaluate a pruned decision tree model that assigns seaweed samples to one of two procurement classes: worthy or not feasible. Model performance was evaluated using six pruning methods: threshold, cost complexity, reduced error, pessimistic error, critical value, and minimum number pruning. Among these, threshold and cost complexity pruning produced the highest classification accuracy at 63.4%, while maintaining model interpretability and minimizing overfitting. The most influential attributes were drying time, moisture content, and price, while the remaining features had a negligible impact. Validation was conducted through bootstrapping, confirming model robustness across sampling variations. The final model was implemented in a web-based interface using explainable artificial intelligence to support real-time, transparent decision-making for buyers and supply chain stakeholders. Despite limitations in the current feature set and the use of a single classifier, the system offers a practical and interpretable tool for quality-based procurement in agri-marine environments. Future research will aim to improve predictive performance by incorporating environmental data, image-based grading, biochemical profiling, and exploring ensemble methods.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100628"},"PeriodicalIF":0.0,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144925150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An interval-valued spherical fuzzy framework for strategic renewable energy selection","authors":"Galip Cihan Yalçın ,&nbsp;Sercan Edinsel ,&nbsp;Prasenjit Chatterjee ,&nbsp;Shervin Zakeri","doi":"10.1016/j.dajour.2025.100625","DOIUrl":"10.1016/j.dajour.2025.100625","url":null,"abstract":"<div><div>Many countries are prioritizing renewable energy sources in response to fossil fuel depletion, environmental concerns, and the need for energy resilience. This study evaluates five renewable energy alternatives: Biomass, Wind, Solar, Geothermal, and Hydro, with the aim of reducing foreign energy dependency and enhancing flexibility under potential geopolitical disruptions. A three-stage hybrid decision-making framework is proposed, integrating Modified Preference Selection Index (MPSI) and Multi-Attributive Border Approximation Area Comparison (MABAC) methods within an Interval-Valued Spherical Fuzzy (IVSF) environment. In the first stage, expert input is collected. The second stage applies IVSF-MPSI to determine the criteria weights under uncertainty. The third stage employs IVSF-MABAC to rank the alternatives based on these weights. The results indicate that Solar Energy, with a distance value of 0.2783, is the most suitable renewable energy, followed by Wind, Hydro, Geothermal, and Biomass. The proposed IVSF-MPSI-MABAC model equips decision-makers with a mathematically rigorous, uncertainty-resilient evaluation framework that supports quantitative trade-off analysis, prioritization of capital-intensive projects, and alignment of renewable energy portfolios with long-term energy security and sustainability objectives, while the integrated sensitivity analysis ensures ranking stability and robustness against variations in decision parameters.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100625"},"PeriodicalIF":0.0,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144988571","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A data-filtered sentiment analysis model for economic forecasting","authors":"Wanwan Zheng ,&nbsp;Kunihiko Hara","doi":"10.1016/j.dajour.2025.100629","DOIUrl":"10.1016/j.dajour.2025.100629","url":null,"abstract":"<div><div>In behavioral economics, sentiments influence decision-making processes, with positive sentiments tending to underestimate risks and negative sentiments overestimate them. At the individual level, these sentiments shape economic behavior in ways that collectively influence broader economic dynamics. With the proliferation of textual data and advancements in natural language processing techniques, the analysis of economic sentiments has garnered growing attention, with large language models showing superior analytical capabilities. However, unlike many machine learning tasks where true labels are available through human annotation, sentiment analysis encounters challenges in obtaining true labels due to the psychological biases and inconsistencies inherent in human assessments. To address this issue, this study introduced a data filtering methodology to enhance data reliability and developed a robust sentiment analysis model tailored to the Japanese economy. The findings revealed that our model not only outperforms existing models in terms of generalization capability across diverse datasets — achieving RMSE values of 0.09–0.11 and classification accuracies of 0.83–0.88 — but also effectively captures fluctuations in other quantitative economic indicators, as evidenced by Euclidean distances of up to 1.56, which is smaller than the records 4.33 and 4.24 of the existing models. Moreover, a statistically significant correlation between qualitative and quantitative economic indicators was identified, highlighting the potential of qualitative indicators in predicting economic conditions.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100629"},"PeriodicalIF":0.0,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144932271","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An adaptive clustering framework for personality prediction using enhanced seed optimization","authors":"Hartono ,&nbsp;Muhammad Khahfi Zuhanda ,&nbsp;Rahmad Syah ,&nbsp;Fikriyah Iftinan Fauzi ,&nbsp;Nini Sri Wahyuni ,&nbsp;Istiana ,&nbsp;Eva Yulina","doi":"10.1016/j.dajour.2025.100630","DOIUrl":"10.1016/j.dajour.2025.100630","url":null,"abstract":"<div><div>Personality prediction has become an increasingly important area in psychological computing and human-centered AI, especially with the rise of user-generated textual data from social media platforms. However, current approaches – primarily based on supervised learning – face major challenges in dealing with class imbalance, noisy inputs, and poor generalization in real-world scenarios. This study introduces an adaptive hybrid clustering framework for MBTI-based personality prediction by integrating K-Means with Nearest Neighbor Density Peak (K-NNDP) and Determinantal Point Process (DPP) to enhance seed optimization. The framework addresses key limitations of traditional clustering methods – such as poor class imbalance handling, lack of diversity, and outlier sensitivity – by combining density-based refinement with probabilistic, diversity-driven seed selection. Applied to the MBTI Kaggle dataset of 8,675 instances, the model transforms unstructured text into numerical vectors using TF-IDF, Bag-of-Words, and GloVe embeddings. Experimental results show that the proposed method outperforms six established supervised models – Decision Trees, KNN, Logistic Regression, LSVC, SGD, and XGBoost – across all multi-label classification metrics, achieving the highest Exact Match Ratio (0.813), Accuracy (0.915), Precision (0.878), Recall (0.897), and F1-Score (0.887), while significantly reducing Hamming Loss (0.103) and Zero-One Loss (0.187). Sensitivity analyses under varying imbalance ratios (up to 1:20), increasing textual noise, and data diversity levels further validate the model’s robustness and generalizability, even in challenging conditions. These findings confirm the effectiveness of the proposed unsupervised approach in uncovering coherent personality clusters without requiring labeled data. Nonetheless, further improvements are needed in enhancing cluster interpretability and optimizing runtime performance. Future research will explore real-time implementation and integration into personality-aware systems</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100630"},"PeriodicalIF":0.0,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145019083","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An adaptive analytics framework for customer retention through integrative feature optimization and ensemble learning","authors":"Rahmad B.Y. Syah ,&nbsp;Marischa Elveny","doi":"10.1016/j.dajour.2025.100626","DOIUrl":"10.1016/j.dajour.2025.100626","url":null,"abstract":"<div><div>An adaptive analytics workflow is presented for customer churn prediction, combining Principal Component Analysis for dimensionality reduction, a hybrid Modified Particle Swarm Gravitational Search Optimization (MPSO-GSO) for feature selection and hyperparameter tuning, and an ensemble learning stage combining XGBoost and LightGBM through weighted voting. Applied to an e-commerce dataset, the complete framework achieves AUC = 0.99 and accuracy = 0.98, outperforming standalone XGBoost (AUC = 0.98) and LightGBM (AUC = 0.97). Stratified 5-fold cross-validation and paired t-tests confirm the statistical significance of this improvement (p &lt; 0.01). Subsequent SHAP analysis interprets the feature contributions, demonstrating that this integrative, optimization-based approach substantially improves the quality of churn prediction.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100626"},"PeriodicalIF":0.0,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144925151","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A closed-form solution approach for optimal reorder point in economic order quantity models with uncertain demands","authors":"Omid Jadidi ,&nbsp;Fatemeh Firouzi ,&nbsp;Shahryar Sorooshian","doi":"10.1016/j.dajour.2025.100622","DOIUrl":"10.1016/j.dajour.2025.100622","url":null,"abstract":"<div><div>The reorder point formula in the Economic Order Quantity (EOQ) model traditionally assumes that demand during the lead-time follows a normal distribution. However, this assumption presents several challenges. First, alternative probability distributions may better capture demand patterns for specific products and markets. Second, historical data may not always be available to predict these distributions; in such cases, fuzzy set theory can be used to estimate demand based on expert opinions and judgments. Third, the conventional reorder point formula overlooks important factors, such as unit wholesale price and unit shortage costs. For instance, when unit shortage or goodwill costs are high, increasing the reorder point can help minimize the risk of stockouts. To address these issues, we reformulate the inventory problem during the lead-time as a newsvendor problem and derive closed-form solutions for the optimal reorder point. In this model, demand during the lead-time is represented using both fuzzy numbers (to capture possibility) and probability distributions, allowing us to incorporate factors like unit wholesale price and shortage or goodwill costs. Additionally, we provide managerial insights through numerical analysis, helping to guide decisions on reorder point adjustments.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100622"},"PeriodicalIF":0.0,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144890673","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An analytical framework for optimizing urban rail schedules with energy recovery and sensor integration","authors":"Hassan Farshad","doi":"10.1016/j.dajour.2025.100623","DOIUrl":"10.1016/j.dajour.2025.100623","url":null,"abstract":"<div><div>Urban railway systems are crucial components of sustainable public transportation but face significant operational costs due to energy consumption and maintenance needs. This study develops a novel optimization framework that integrates regenerative braking strategies and Internet of Things (IoT) adoption into train scheduling for improved energy efficiency and reliability. A real-world case study was conducted using data from Iran Urban Railway Organization. The proposed model was solved using CPLEX in GAMS software and tested under various adoption rates of IoT technologies. Results demonstrate that an optimal IoT adoption rate of 0.7 minimizes total operational cost, achieving a cost reduction from 1,687,600 (at 0 adoption) to 1,265,432 units. This rate also balances implementation cost (4,682,356 units) and leads to a 52% decrease in quality-related costs. Moreover, train schedule optimization improved timing consistency: dwell times were stabilized at 1.5–2 min, with longer stops (5 min) at major stations, and train speeds ranged between 30–43 km/h. These improvements enhance service reliability and enable significant energy recovery through regenerative braking. This research provides a robust decision-support tool for railway operators by combining IoT-based predictive maintenance and energy-aware train scheduling, offering measurable cost and performance benefits in real-world operations.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100623"},"PeriodicalIF":0.0,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144890675","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An analytics approach to extracting location and sentiment insights from classified social media data","authors":"Fernando Lovera ,&nbsp;Yudith Cardinale","doi":"10.1016/j.dajour.2025.100624","DOIUrl":"10.1016/j.dajour.2025.100624","url":null,"abstract":"<div><div>Social networks are becoming vital for people to interact with each other, which in turn represent the production of a huge amount of information that can be useful in many contexts (e.g., medicine, natural disasters, commercial purposes, tourism). Nevertheless, analyzing such Big Data for insights might be difficult if the appropriate processing and analysis tools are not available. In this sense, we propose a framework to analyze social network content by integrating Topic Detection, Sentiment Analysis, and Geolocation. The gathered information is processed using Natural Language Processing methods to extract textual elements that make it possible for each framework component to function as intended. After reading through a stream of posts, the Topic Detection method classifies them and removes any that have nothing to do with the subject being analyzed. Sentiment Analysis component combines Machine Learning, Knowledge Graphs, and Semantic Web techniques, using SPARQL in conjunction with DBpedia and Nominatim. The Geolocation component scans posts and attempts to determine their geographical position. In this study, we implement a proof-of-concept on X (formerly Twitter), called XAF (X Analyzer Framework), to work in the context of natural disasters, to show the efficiency of combining Sentiment Analysis, Geolocation, and Topic Detection, and the possibility to be used in other contexts. We describe the general architecture of XAF and show the performance of each module as well as the holistic solution. Results show that XAF provides a platform to analyze X posts from different perspectives that allows implementing applications able to respond in real time.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100624"},"PeriodicalIF":0.0,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144890674","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An evolutionary analytics model for enhancing food distribution through two-tier routing","authors":"Muhammad Khahfi Zuhanda ,&nbsp;Hartono ,&nbsp;Samsul A. Rahman Sidik Hasibuan ,&nbsp;Najwa Muthmainnah ,&nbsp;Marlina Br Girsang","doi":"10.1016/j.dajour.2025.100621","DOIUrl":"10.1016/j.dajour.2025.100621","url":null,"abstract":"<div><div>Food insecurity remains a global challenge that demands efficient and equitable logistical solutions, especially in the distribution of food aid. This study addresses the Two-Echelon Vehicle Routing Problem (2E-VRP) in foodbank logistics by proposing a novel multi-objective optimization framework. The model combines the Non-dominated Sorting Genetic Algorithm II (NSGA-II) with k-means clustering, a rare approach in food distribution logistics, to optimize route efficiency. The first echelon involves transporting food from a central depot to intermediate agents using trucks, while the second echelon involves delivering food from agents to beneficiaries using vans. The primary objectives are to minimize total delivery distance and reduce delivery time, while considering fleet capacities, time window constraints, and ensuring fair distribution. Using spatial and demand data from Medan, Indonesia, the study evaluates the model’s performance, computational efficiency, and sensitivity to logistical factors such as fleet size. Results show that clustering improves route compactness and reduces travel distance, especially in large-scale networks. However, it increases computational time, highlighting a trade-off between solution quality and complexity. This research offers a scalable, data-driven framework for foodbank logistics, contributing to sustainable logistics and providing an innovative solution for optimizing food distribution in both urban and rural settings.</div></div>","PeriodicalId":100357,"journal":{"name":"Decision Analytics Journal","volume":"16 ","pages":"Article 100621"},"PeriodicalIF":0.0,"publicationDate":"2025-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144893394","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}