Human-machine collaboration for feature selection and integration to improve congestive Heart failure risk prediction

2023 | Decision Support Systems | Citations: 0

Authors: Ben-Assuli, Ofir; Heart, Tsipi; Klempfner, Robert; Padman, Rema

Abstract: The issue of harnessing machine learning (ML) algorithms for the prediction of a ... Expand

Abstract: The issue of harnessing machine learning (ML) algorithms for the prediction of adverse medical events is important considering the prevalence of vast repositories of patient-level medical data, and the quest to reduce healthcare costs while elevating the quality of care. Of particular interest is identifying a set of features that can better predict the risk of unplanned readmissions of chronically ill patients within 30 days post discharge, among them patients with Congestive Heart Failure (CHF). While numerous studies have employed ML algorithms to identify sets of significant features, only a handful have compared features extracted by various methods, and even fewer have compared the predictive power of feature sets selected by human experts to those selected by ML. Hence, this research aimed to elicit a comprehensive feature set selected by four ML algorithms and compare its predictive performance to that of a feature set selected by experts. We then evaluated a merged list of the two feature sets, constructing a human-machine collaborative set to predict the likelihood of 30-day unplanned readmission of CHF patients using data on 10,763 CHF patients hospitalized during the years 2010–2017. Our models achieved Area under the ROC Curve (AUC) above 0.8 and an Accuracy of ~0.89, comparable to the best models in the extant literature. Moreover, the Human-Machine Collaborative model significantly outperformed (best AUC 0.836) other human-only selected models as well as the machine-selected set. This study contributes to a better understanding of the power of using ML for patient risk stratification with special attention to the benefits of human-machine collaboration, even when these two entities work separately or in parallel, for improved clinical decision making. Collapse

Topics: healthcare data decision making big data missing data data warehouse

Methods: machine learning computational algorithm XGBoost descriptive statistic random forest classification

Trust in Online Ride-Sharing Transactions: Impacts of Heterogeneous Order Features

2023 | Journal of Management Information Systems | Citations: 0

Authors: Cheng, Xusen; Fu, Shixuan; Sun, Jianshan; Zuo, Meiyun; Meng, Xiangsong

Abstract: With the development of the sharing economy, online ride-sharing has become a pr ... Expand

Abstract: With the development of the sharing economy, online ride-sharing has become a primary form of commuting. Using secondary transaction data, this study investigates the associations between the heterogeneous features and mutual trust in sharing economy-driven online ride-sharing transactions. Based on an examination of 12,404 ride-sharing orders in Beijing, we propose a set of trust distribution maps using order location data to reveal heterogeneous spatial patterns of the relationship between online ride-sharing transactions and mutual trust. The results show that the historical order completion rate and order distance are positively associated with mutual trust in ride-sharing transactions, whereas order time and departure density negatively and significantly influence mutual trust. Furthermore, we use machine learning algorithms to predict trust. The implications for theory and practice and future research directions are discussed. Collapse

Topics: ride sharing platform sharing platform logistics management geographic information system online review

Methods: machine learning computational algorithm regression analysis method experiment exploratory data analysis

Theories: signaling theory

Social Determinants of Health and ER Utilization: Role of Information Integration during COVID-19

2023 | ACM Transactions on Management Information Systems | Citations: 0

Authors: Guo, Tianjian; Bardhan, Indranil; Khurshid, Anjum

Abstract: Emergency room (ER) admissions are the front door for the utilization of a commu ... Expand

Abstract: Emergency room (ER) admissions are the front door for the utilization of a community's health resources and serve as a valuable proxy for a community health system's capacity. While recent research suggests that social determinants of health (SDOH) are important predictors of patient health outcomes, their impact on ER utilization during the COVID-19 pandemic is not well understood. Further, the role of hospital information integration in moderating the impact of SDOH on ER utilization has not received adequate attention. Utilizing longitudinal claims data from a regional health information exchange spanning 6 years including the COVID-19 period, we study how SDOH affects ER utilization and whether effective integration of patient health information across hospitals can moderate its impact. Our results suggest that a patient's economic well-being significantly reduces future ER utilization. The magnitude of this relationship is significant when patients are treated at hospitals with high information integration but is weaker when patients receive care at hospitals with lower levels of information integration. Instead, patients' family and social support can reduce ER utilization when they are treated at hospitals with low information integration. In other words, different dimensions of SDOH are important in low versus high information integration conditions. Furthermore, predictive modeling shows that patient visit type and prior visit history can significantly improve the predictive accuracy of ER utilization. Our research implications support efforts to develop national standards for the collection and sharing of SDOH data and their use and interpretation for clinical decision making by healthcare providers and policy makers. Collapse

Topics: healthcare data pandemic information exchange accounting health information system

Methods: experimental group machine learning XGBoost longitudinal research econometric modeling

Responsible Artificial Intelligence in Healthcare: Predicting and Preventing Insurance Claim Denials for Economic and Social Wellbeing

2023 | Information Systems Frontiers | Citations: 1

Authors: Johnson, Marina; Albizri, Abdullah; Harfouche, Antoine

Abstract: It is estimated that one out of seven health insurance claims is rejected in the ... Expand

Abstract: It is estimated that one out of seven health insurance claims is rejected in the US; hospitals across the country lose approximately $262 billion annually due to denied claims. This widespread problem causes huge cash-flow issues and overburdens patients. Thus, preventing claim denials before claims are submitted to insurers improves profitability, accelerates the revenue cycle, and supports patients’ wellbeing. This study utilizes Design Science Research (DSR) paradigm and develops a Responsible Artificial Intelligence (RAI) solution helping hospital administrators identify potentially denied claims. Guided by five principles, this framework utilizes six AI algorithms – classified as white-box and glass-box – and employs cross-validation to tune hyperparameters and determine the best model. The results show that a white-box algorithm (AdaBoost) model yields an AUC rate of 0.83, outperforming all other models. This research’s primary implications are to (1) help providers reduce operational costs and increase the efficiency of insurance claim processes (2) help patients focus on their recovery instead of dealing with appealing claims. Collapse

Topics: artificial intelligence ADA missing data partial dependence plot responsible artificial intelligence

Methods: machine learning computational algorithm experiment support vector machine design artifact

Diffusion of real versus misinformation during a crisis event: A big data-driven approach

2023 | International Journal of Information Management | Citations: 0

Authors: King, Kelvin K.; Wang, Bin

Abstract: Misinformation has captured the interest of academia in recent years with severa ... Expand

Abstract: Misinformation has captured the interest of academia in recent years with several studies looking at the topic broadly with inconsistent results. In this research, we attempt to bridge the gap in the literature by examining the impacts of user-, time-, and content-based characteristics that affect the virality of real versus misinformation during a crisis event. Using a big data-driven approach, we collected over 42 million tweets during Hurricane Harvey and obtained 3589 original verified real or false tweets by cross-checking with fact-checking websites and a relevant federal agency. Our results show that virality is higher for misinformation, novel tweets, and tweets with negative sentiment or lower lexical density. In addition, we reveal the opposite impacts of sentiment on the virality of real news versus misinformation. We also find that tweets on the environment are less likely to go viral than the baseline religious news, while real social news tweets are more likely to go viral than misin­ formation on social news. Collapse

Topics: Twitter misinformation social media hurricane big data

Methods: machine learning computational algorithm natural language processing decision tree classification gradient boosting

ROLEX: A Novel Method for Interpretable Machine Learning Using Robust Local Explanations

2023 | Management Information Systems Quarterly | Citations: 2

Authors: Kim, Buomsoo (Raymond); Srinivasan, Karthik; Kong, Sung Hye; Kim, Jung Hee; Shin, Chan Soo; Ram, Sudha

Abstract: Recent developments in big data technologies are revolutionizing the field of h ... Expand

Abstract: Recent developments in big data technologies are revolutionizing the field of healthcare predictive analytics (HPA), enabling researchers to explore challenging problems using complex prediction models. Nevertheless, healthcare practitioners are reluctant to adopt those models as they are less transparent and accountable due to their black-box structure. We believe that instance-level, or local, explanations enhance patient safety and foster trust by enabling patient-level interpretations and medical knowledge discovery. Therefore, we propose the RObust Local EXplanations (ROLEX) method to develop robust, instance-level explanations for HPA models in this study. ROLEX adapts state-of-the-art methods and ameliorates their shortcomings in explaining individual-level predictions made by black-box machine learning models. Our analysis with a large real-world dataset related to a prevalent medical condition called fragility fracture and two publicly available healthcare datasets reveals that ROLEX outperforms widely accepted benchmark methods in terms of local faithfulness of explanations. In addition, ROLEX is more robust since it does not rely on extensive hyperparameter tuning or heuristic algorithms. Explanations generated by ROLEX, along with the prototype user interface presented in this study, have the potential to promote personalized care and precision medicine by providing patient-level interpretations and novel insights. We discuss the theoretical implications of our study in healthcare, big data, and design science. Collapse

Topics: explainable artificial intelligence usability user interface local interpretable model agnostic explanation problem solving

Methods: machine learning design artifact computational algorithm longitudinal research ablation analysis

Theories: lemon market theory

Towards Adversarially Superior Malware Detection Models: An Adversary Aware Proactive Approach using Adversarial Attacks and Defenses

2023 | Information Systems Frontiers | Citations: 1

Authors: Rathore, Hemant; Samavedhi, Adithya; Sahay, Sanjay K.; Sewak, Mohit

Abstract: The android ecosystem (smartphones, tablets, etc.) has grown manifold in the las ... Expand

Abstract: The android ecosystem (smartphones, tablets, etc.) has grown manifold in the last decade. However, the exponential surge of android malware is threatening the ecosystem. Literature suggests that android malware can be detected using machine and deep learning classifiers; however, these detection models might be vulnerable to adversarial attacks. This work investigates the adversarial robustness of twenty-four diverse malware detection models developed using two features and twelve learning algorithms across four categories (machine learning, bagging classifiers, boosting classifiers, and neural network). We stepped into the adversary’s shoes and proposed two false-negative evasion attacks, namely GradAA and GreedAA, to expose vulnerabilities in the above detection models. The evasion attack agents transform malware applications into adversarial malware applications by adding minimum noise (maximum five perturbations) while maintaining the modified applications’ structural, syntactic, and behavioral integrity. These adversarial malware applications force misclassifications and are predicted as benign by the detection models. The evasion attacks achieved an average fooling rate of 83.34% (GradAA) and 99.21% (GreedAA) which reduced the average accuracy from 90.35% to 55.22% (GradAA) and 48.29% (GreedAA) in twenty-four detection models. We also proposed two defense strategies, namely Adversarial Retraining and Correlation Distillation Retraining as countermeasures to protect detection models from adversarial attacks. The defense strategies slightly improved the detection accuracy but drastically enhanced the adversarial robustness of detection models. Finally, investigating the robustness of malware detection models against adversarial attacks is an essential step before their real-world deployment and can help in developing adversarially superior detection models. Collapse

Topics: malware Android mobile application database system Google+

Methods: machine learning artificial neural network computational algorithm deep learning descriptive statistic

Improving Information Systems Sustainability by Applying Machine Learning to Detect and Reduce Data Waste

2023 | Communications of the Association for Information Systems | Citations: 0

Authors: Savarimuthu, Bastin Tony Roy; Corbett, Jacqueline; Yasir, Muhammad; Lakshmi, Vijaya

Abstract: Big data are key building blocks for creating information value. However, infor ... Expand

Abstract: Big data are key building blocks for creating information value. However, information systems are increasingly plagued with useless, waste data that can impede their effective use and threaten sustainability objectives. Using a constructive design science approach, this work first, defines digital data waste. Then, it develops an ensemble artifact comprising two components. The first component comprises 13 machine learning models for detecting data waste. Applying these to 35,576 online reviews in two domains reveals data waste of 1.9% for restaurant reviews compared to 35.8% for app reviews. Machine learning can accurately identify 83% to 99.8% of data waste; deep learning models are particularly promising, with accuracy ranging from 96.4% to 99.8%. The second component comprises a sustainability cost calculator to quantify the social, economic, and environmental benefits of reducing data waste. Eliminating 5948 useless reviews in the sample would result in saving 6.9 person hours, $2.93 in server, middleware and client costs, and 9.52 kg of carbon emissions. Extrapolating these results to reviews on the internet shows substantially greater savings. This work contributes to design knowledge relating to sustainable information systems by highlighting the new class of problem of data waste and by designing approaches for addressing this problem. Collapse

Topics: mobile application online review Python data management Google+

Methods: deep learning product review computational algorithm design artifact RoBERTa

How Explainable AI Methods Support Data-Driven Decision Making

2023 | International Conference on Business Informatics | Citations: 0

Authors: Stoffels, Dominik; Grabl, Susanne; Fischer, Thomas; Fiedler, Marina

Abstract: Explainable AI (XAI) holds great potential to reveal the patterns in black-box ... Expand

Abstract: Explainable AI (XAI) holds great potential to reveal the patterns in black-box AI models and to support data-driven decision-making. We apply four post-hoc explanatory methods to demonstrate the explanatory capabilities of these methods for data-driven decision-making using the illustrative example of unwanted job turnover and human resource management (HRM) support. We show that XAI can be a useful aid in data-driven decision-making, but also highlight potential drawbacks and limitations of which users in research and practice should be aware. Collapse

Topics: explainable artificial intelligence shapley additive explanation decision making local interpretable model agnostic explanation human resource management

Methods: machine learning support vector machine AdaBoost company material computational algorithm

Efficient fraud detection using deep boosting decision trees

2023 | Decision Support Systems | Citations: 0

Authors: Xu, Biao; Wang, Yao; Liao, Xiuwu; Wang, Kaidong

Abstract: Fraud detection is to identify, monitor, and prevent potentially fraudulent acti ... Expand

Abstract: Fraud detection is to identify, monitor, and prevent potentially fraudulent activities from complex data. The recent development and success in AI, especially machine learning, provides a new data-driven way to deal with fraud. From a methodological point of view, machine learning based fraud detection can be divided into two categories, i.e., conventional methods (decision tree, tree boosting methods, etc.) and deep learning, both of which have significant limitations in terms of the lack of representation learning ability for the former and interpretability for the latter. Furthermore, due to the rarity of detected fraud cases, the associated data is usually imbalanced, which seriously degrades the performance of classification algorithms. Collapse

Topics: fraud detection marketing management

Methods: machine learning computational algorithm decision tree classification artificial neural network gradient boosting