Predictive Analytics Approach Research Articles

Plant nutritional deficiency detection: a survey of predictive analytics approaches

Plant nutritional deficiency detection: a survey of predictive analytics approaches

Evaluating the impact of AI and blockchain on credit risk mitigation: A predictive analytic approach using machine learning

In recent years, the integration of artificial intelligence (AI), blockchain technology, and machine learning has transformed credit risk mitigation strategies in the financial industry. This paper explores the practical applications of these technologies in identifying, assessing, and managing credit risk, with a specific focus on predictive analytics and decentralized frameworks. Through a comprehensive literature review and case studies, the research demonstrates how AI-driven algorithms, blockchain's transparent and immutable ledger systems, and machine learning models have enhanced the precision and efficiency of credit risk evaluations. Additionally, the study investigates how these innovations are being adopted by financial institutions to create more accurate credit scoring systems, reduce fraud, and optimize operational risk management. While these technologies hold great promise, challenges such as data privacy, regulatory compliance, and implementation costs remain significant barriers. The paper concludes with recommendations for overcoming these challenges and maximizing the potential of AI, blockchain, and machine learning in credit risk mitigation.

Impact of IoT on Business Decision-Making: A Predictive Analytics Approach

To illustrate, the Internet of Things (IoT) has done an excellent job of changing the way business organizational decisions are made by making data available in unimaginable magnitudes. The following paper aims to identify and explain how IoT, with the help of predictive analytics, can improve BI and decision-making. This includes a discussion on several operational and customer cases, consideration of Artificial Intelligence and Machine Learning applications alongside the Internet of Things, and analysis of trends and developments of IoT. These are the core conclusions of IoT importance within real-time business, and related difficulties and tools for the improvement.

PREDICTING FLIGHT DELAYS WITH ERROR CALCULATION USING MACHINE LEARNED CLASSIFIERS

Worldwide, flight delays are becoming a major issue for the airline business. Due to air traffic congestion brought on by the airline industry's expansion over the last 20 years, flights have been delayed. In addition to hurting the economy, flight delays also have a detrimental effect on the environment since they increase fuel consumption and gas emissions. Thus, it is essential to take all reasonable precautions to avoid flight delays and cancellations. This paper's primary goal is to forecast an airline's delay utilizing a variety of variables. Thus, it is necessary to perform forward-looking analysis, which encompasses a variety of algorithmic predictive analytics approaches that use historical and current data to create models that are used for forecasts or simply to look at future delays using machine learning algorithms like Python 3's Gradient Boosting Regression technique, Bayesian Ridge, Random Forest Regression, and Logistic Regression. This will make it easier for the user to forecast whether an aircraft will arrive on time or not. Additionally, delay prediction analysis will assist airline industries in reducing their losses.

An In-Depth Review of Predictive Methods for Oil and Gas Applications

The oil and gas industry aims to optimize production, reduce costs, and increase efficiency. Predictive models have gained popularity as potential solutions in recent years. In predictive models, machine learning algorithms analyse oil and gas operations data and forecast future performance. This paper examines the current state of predictive modelling in the oil and gas industry with the objective to systematically review and analyse current research on the predictive modelling in the oil and gas industry. The paper begins by highlighting the sub-fields and datasets in the oil and gas industry that used recent machine learning methods for predictive modelling. Additionally, literature from the Scopus and Web of Science indexes was reviewed. This study assessed recent approaches for oil and gas industry in predictive applications for papers published up until December 2022. The findings identify several advantages and disadvantages that can be used as guidelines to effectively implement predictive modelling in the oil and gas industry. It includes challenges on the requirement for accurate and reliable data, the development of appropriate algorithms, and the integration of predictive models into existing workflows. In addition, the finding highlights the growing application of deep learning algorithms for various tasks as one of the major trends. From the analysis of the state-of-the-art in predictive techniques, it is necessary first to survey the landscape of existing predictive analytics approaches and their methods to employ in oil and gas prediction.

Prediction of water–main failures and management of the associated risks using integrated predictive analytics approach

Water distribution networks are vital to delivering potable water. They commonly include interconnected water mains (WMs), pumps and other hydraulic controls. Numerous WM failures have occurred due to ageing and harsh climate. This paper aims to predict the probability of future failures and integrate the predictions into risk-management strategies. The novelty lies in emphasising on the relevance to networks in cold regions like Canada. This study applied clustering and principal component analysis to the WM data from the Canadian City of Kitchener network. Clustering is shown to improve the failure prediction outcomes from the random forest algorithm and risk analysis output. Compared to without implementing clustering, the improvement reached 67–80% for WMs with high-rating risk. This paper successfully produced risk maps for Kitchener’s network, showing that only a small percentage (0.07–1.02%) of the existing WMs needs immediate action (prioritised rehabilitation or replacement). In addition to WM length and diameter, freeze index is shown to be an influence factor for failure predictions. The integrated, proactive approach discussed in this article can be applied to other cold-region WDNs. The results help reduce water losses and develop cost-effective, practical risk-management strategies.

Sustainability with Limited Data: A Novel Predictive Analytics Approach for Forecasting CO2 Emissions

AbstractUnforeseen events (e.g., COVID-19, the Russia-Ukraine conflict) create significant challenges for accurately predicting CO2 emissions in the airline industry. These events severely disrupt air travel by grounding planes and creating unpredictable, ad hoc flight schedules. This leads to many missing data points and data quality issues in the emission datasets, hampering accurate prediction. To address this issue, we develop a predictive analytics method to forecast CO2 emissions using a unique dataset of monthly emissions from 29,707 aircraft. Our approach outperforms prominent machine learning techniques in both accuracy and computational time. This paper contributes to theoretical knowledge in three ways: 1) advancing predictive analytics theory, 2) illustrating the organisational benefits of using analytics for decision-making, and 3) contributing to the growing focus on aviation in information systems literature. From a practical standpoint, our industry partner adopted our forecasting approach under an evaluation licence into their client-facing CO2 emissions platform.

A predictive analytics approach for forecasting bike rental demand

The demand for rental bikes in urban areas fluctuates, leading to localized surpluses and shortages. To address this challenge, effective bike relocation strategies are essential for ensuring equitable distribution and maximizing customer satisfaction. This study aims to employ advanced machine learning techniques to forecast bike rental demand in urban areas, thereby enhancing the efficiency and accessibility of bike rental services and contributing to sustainable urban mobility. The study comprehensively analyzes various influencing factors using machine learning models, including Ordinary Least Squares regression, MLP Regression, Gradient Boosting Regression, Random Forest Regression, Polynomial Regression, and Decision Tree Regression. The primary objective is to identify the most accurate predictor by comparing key metrics such as R-squared (R2), Mean Squared Error (MSE), Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Pearson Correlation Coefficient. Insights gained from this analysis aid in identifying influential variables and ensure the development of resource-efficient and adaptable models, leading to more informed decision-making for rental bike businesses. Additionally, future research directions involve the implementation of artificial intelligence technology to predict overall bike demand based on urban cities’ criteria, including the number of national and international tourists. By addressing these objectives, this study seeks to provide valuable insights and tools for rental bike businesses to optimize operations, make strategic decisions, and enhance customer experience in competitive urban markets.

Coupling simulation and machine learning for predictive analytics in supply chain management

Predictive analytics is the approach to business analytics that answers the question of what might happen in the future. Although predictive information is critical for making forward-looking decisions, traditional approaches struggle to cope with the increasing uncertainty and complexity that characterise modern supply chains. Simulation is limited by insufficient timeliness, while machine learning is constrained by poor interpretability and data scarcity. Inspired by the complementary nature of simulation and machine learning, an integrated predictive analytics approach is proposed and applied to a humanitarian supply chain. By coupling simulation and machine learning, predictive models can be developed with limited historical data, and pre-crisis performance assessment can be performed to facilitate timely and informed decisions. The proposed approach enables managers to gain valuable insights into the complex evolution of the uncertain future, which also opens up the possibility of further integration with optimisation and digital twins.

A predictive approach for myocardial infarction risk assessment using machine learning and big clinical data

Myocardial infarction is one of the most common cardiovascular diseases in emergency departments. Early prevention of this dangerous condition significantly impacts public health and considerable socioeconomic outcomes. The emergence of electronic health records (EHR) and the availability of real-world clinical data have provided opportunities to improve the quality and efficiency of healthcare by using artificial intelligence tools. In this study, we focus on the early recognition of risk factors, which can provide valuable information for early prediction of myocardial infarction and promoting a healthy life. Based on a big clinical dataset, we develop a predictive analytics approach for myocardial infarction. A vital step in efficient prediction is assessing the significance of input features, their relationships and their contributions to the disease. Therefore, we adopted statistical techniques, principal component analysis (PCA) and feature engineering. To reveal patterns and insights on our dataset, we implemented machine learning (ML) models varying from classical to more sophisticated: decision trees (DT), random forests (RF), gradient boosting algorithms (GBoost, LightGBM, CatBoost, and XGBoost) and deep neural networks (DNN). The imbalance-data issue is tackled by employing random under-sampling technique. The light gradient boosting model (LightGBM) with feature engineering on the balanced dataset is the best prediction performance achieved in this study.

Deconstructing the Normalization of Data Colonialism in Educational Technology

As learning analytics and educational data mining have become the “new normal” in the field, scholars have observed the emergence of data colonialism. Generally, data colonialism can be understood as the process by which data were considered “free” to take and appropriate. Building on this theoretical understanding, this study aims to contextualize data colonialism in educational technology by identifying and reviewing learning analytics studies that adopted a predictive analytics approach. We examined 22 studies from major educational technology journals and noted how they (1) see data as a resource to appropriate, (2) establish new social relations, (3) show the concentration of wealth, and (4) promote ideologies. We found evidence of data colonialism in the field of educational technology. While these studies may promote “better” ideologies, it is concerning how they justify the authorities capitalizing on “free” data. After providing a contextualized view of data colonialism in educational technology, we propose several measures to decolonialize data practices, adopting a postcolonialist approach. We see data colonialism not only as a privacy issue but also as a culture that must be challenged.

Machine Learning Predictions on Outpatient No-Show Appointments in a Malaysia Major Tertiary Hospital.

A no-show appointment occurs when a patient does not attend a previously booked appointment. This situation can cause other problems, such as discontinuity of patient treatments as well as a waste of both human and financial resources. One of the latest approaches to address this issue is predicting no-shows using machine learning techniques. This study aims to propose a predictive analytical approach for developing a patient no-show appointment model in Hospital Kuala Lumpur (HKL) using machine learning algorithms. This study uses outpatient data from the HKL's Patient Management System (SPP) throughout 2019. The final data set has 246,943 appointment records with 13 attributes used for both descriptive and predictive analyses. The predictive analysis was carried out using seven machine learning algorithms, namely, logistic regression (LR), decision tree (DT), k-near neighbours (k-NN), Naïve Bayes (NB), random forest (RF), gradient boosting (GB) and multilayer perceptron (MLP). The descriptive analysis showed that the no-show rate was 28%, and attributes such as the month of the appointment and the gender of the patient seem to influence the possibility of a patient not showing up. Evaluation of the predictive model found that the GB model had the highest accuracy of 78%, F1 score of 0.76 and area under the curve (AUC) value of 0.65. The predictive model could be used to formulate intervention steps to reduce no-shows, improving patient care quality.

Studying infant mortality: A demographic analysis based on data mining models.

Child mortality, particularly among infants below 5 years, is a significant community well-being concern worldwide. The health sector's top priority in emerging states is to minimize children's death and enhance infant health. Despite a substantial decrease in worldwide deaths of children below 5 years, it remains a significant community well-being concern. Children under five years of age died at 37 per 1,000 live birth globally in 2020. However, in underdeveloped countries such as Pakistan and Ethiopia, the fatality rate of children per 1,000 live birth is 65.2 and 48.7, respectively, making it challenging to reduce. Predictive analytics approaches have become well-known for predicting future trends based on previous data and extracting meaningful patterns and connections between parameters in the healthcare industry. As a result, the objective of this study was to use data mining techniques to categorize and highlight the important causes of infant death. Datasets from the Pakistan Demographic Health Survey and the Ethiopian Demographic Health Survey revealed key characteristics in terms of factors that influence child mortality. A total of 12,654 and 12,869 records from both datasets were examined using the Bayesian network, tree (J-48), rule induction (PART), random forest, and multi-level perceptron techniques. On both datasets, various techniques were evaluated with the aforementioned classifiers. The best average accuracy of 97.8% was achieved by the best model, which forecasts the frequency of child deaths. This model can therefore estimate the mortality rates of children under five years in Ethiopia and Pakistan. Therefore, an online model to forecast child death based on our research is urgently needed and will be a useful intervention in healthcare.

Artificial neural network approaches to identify maternal and infant risk and asset factors using Peridata.Net: a WI-MIOS study.

To analyze PeriData.Net, a clinical registry with linked maternal-infant hospital data of Milwaukee County residents, to demonstrate a predictive analytic approach to perinatal infant risk assessment. Using unsupervised learning, we identified infant birth clusters with similar multivariate health indicator patterns, measured using perinatal variables from 2008 to 2019 from n = 43 969 clinical registry records in Milwaukee County, WI, followed by supervised learning risk-propagation modeling to identify key maternal factors. To understand the relationship between socioeconomic status (SES) and birth outcome cluster assignment, we recoded zip codes in Peridata.Net according to SES level. Three self-organizing map clusters describe infant birth outcome patterns that are similar in the multivariate space. Birth outcome clusters showed higher hazard birth outcome patterns in cluster 3 than clusters 1 and 2. Cluster 3 was associated with lower Apgar scores at 1 and 5 min after birth, shorter infant length, and premature birth. Prediction profiles of birth clusters indicate the most sensitivity to pregnancy weight loss and prenatal visits. Majority of infants assigned to cluster 3 were in the 2 lowest SES levels. Using an extensive perinatal clinical registry, we found that the strongest predictive performance, when considering cluster membership using supervised learning, was achieved by incorporating social and behavioral risk factors. There were inequalities in infant birth outcomes based on SES. Identifying infant risk hazard profiles can contribute to knowledge discovery and guide future research directions. Additionally, presenting the results to community members can build consensus for community-identified health and risk indicator prioritization for intervention development.

A strategic roadmap to overcome blockchain technology barriers for sustainable construction: A deep learning-based dual-stage SEM-ANN approach

The adoption of blockchain technology in the sustainable construction industry in India is slow. Existing literature in this area has primarily focused on the potential applications of blockchain technology in construction but little attention is given to the barriers, impeding its adoption. This study fills this research gap by identifying the most significant barriers to adopting blockchain technology for India's sustainable construction industry. The collected data were analyzed via a two-stage PLS-SEM-artificial-neural-network (ANN) predictive analytical approach. In total, 722 construction stakeholder surveys were conducted, and the final model of barriers to adopting blockchain technology for sustainable construction was statistically validated. This study's findings suggest that significant organizational, technological, cultural, legal, security and government barriers limit the adoption of blockchain technology for sustainable construction in India. These findings also imply that effective legislative processes and economic incentives are crucial for ensuring blockchain technology integration into sustainable construction and for efficient implementation of practices. As a further step, a roadmap is developed to support decision-makers in overcoming these barriers in the short, medium, and long terms. Finally, multi-tiered strategies that construction mapping, sustainability, and integration should be adopted to ensure the successful integration of BT into sustainable construction practices in India.

A predictive decision analytics approach for primary care operations management: A case study of double-booking strategy design and evaluation

Primary care plays a vital role for individuals and families in accessing care, staying well, and improving quality of life. However, the complexities and uncertainties in the primary care delivery system (e.g., patient no-shows/walk-ins, staffing shortage) have brought significant challenges in its operations management, which can potentially lead to poor patient outcomes and negative primary care operations (e.g., loss of productivity, inefficiency). This paper presents a decision analytics approach developed based on predictive analytics and simulation modeling to better facilitate management of the underlying complexities and uncertainties in primary care operations. A case study was conducted in a local family medicine clinic to demonstrate the use of this approach to manage patient no-shows. In this case study, a patient no-show prediction model was used in conjunction with an integrated agent-based and discrete-event simulation model to design and evaluate double-booking strategies. Using the predicted patient no-show information, a prediction-based double-booking strategy was created and compared against two other strategies, namely random and designated time. Scenario-based experiments were then conducted to examine the impacts of different double-booking strategies on clinic’s operational outcomes, focusing on the trade-offs between the clinic productivity (measured by daily patient throughput) and efficiency (measured by visit cycle and patient wait time for doctor). The results showed that the best productivity-efficiency balance was derived under the prediction-based double-booking strategy. The proposed hybrid decision analytics approach has the potential to better support decision-making in primary care operations management and improve the system’s performance. Further, it can be generalized in the context of various healthcare settings for broader applications.

A developer-oriented recommender model for the app store: A predictive network analytics approach

While thousands of new mobile applications (i.e., apps) are being added to the major app markets daily, only a small portion of them attain their financial goals and survive in these competitive marketplaces. A key to the quick growth and success of relatively less popular apps is that they should make their way to the limited list of apps recommended to users of already popular apps; however, the focus of the current literature on consumers has created a void of design principles for app developers. In this study, employing a predictive network analytics approach combined with deep learning-based natural language processing and explainable artificial intelligence techniques, we shift the focus from consumers and propose a developer-oriented recommender model. We employ a set of app-specific and network-driven variables to present a novel approach for predicting potential recommendation relationships among apps, which enables app developers and marketers to characterize and target appropriate consumers. We validate the proposed model using a large (>23,000), longitudinal dataset of medical apps collected from the iOS App Store at two time points. From a total of 10,234 network links (recommendations) formed between the two data collection points, the proposed approach was able to correctly predict 8,780 links (i.e., 85.8 %). We perform Shapley Additive exPlanation (SHAP) analysis to identify the most important determinants of link formations and provide insights for the app developers about the factors and design principles they can incorporate into their development process to maximize the chances of success for their apps.

Poverty and distribution of wealth in the USA: a descriptive and predictive analytics approach

Poverty and distribution of wealth in the USA: a descriptive and predictive analytics approach

Poverty and distribution of wealth in the USA-a descriptive and predictive analytics approach

Poverty and distribution of wealth in the USA-a descriptive and predictive analytics approach

Predictive Analytics of Mental Health Problems Among College Students in the Covid-19 Pandemic

Covid-19 pandemic has brought into focus the mental health of various segments of society. This paper presents an empirical study on mental health of college students due to the specific situations caused by the pandemic. A particular focus is on the effects of three stressors, namely, academic workload, separation from school, and fears of contagion among the college students. This initiative was a follow-up of the study by Yang et al. (2021) who proposed a research model that evaluates the impacts of these stressors on perceived stress which subsequently affects the mental health of the students. Using the data collected by Yang et al. (2021), an alternative predictive analytics approach, i.e., Partial Least Squares Structural Equation Modelling (PLS-SEM), was adopted to re-evaluate the research model. This has produced an improvement over the results of Confirmatory Factor Analysis (CFA) adopted by Yang et al. (2021), particularly in the test for the significance of correlation between academic workload and mental health. While the use of PLS-SEM that allowed for strategic refinements of the model produced a significant correlation between the two important constructs, the CFA failed to obtain a similar result. All the other significant correlations between the stressors and mental health and correlation between the mediating factor, the perceived stress, and mental health were also established using the PLS-SEM approach.