Implementing Machine Learning Research Articles

Adaptive learning engine for driving marketing channel performance: a machine learning approach

PurposeThis paper aims to develop and implement a machine learning recommendation engine – an adaptive learning engine that drives business revenue through the ranking and recommendation of offers at a granular customer level across the inbound marketing channels.Design/methodology/approachA data set of over 300,000 unique sample of mobile customers was extracted and prepared. The gradient boosting machine (GBM) algorithm was developed, consolidated, deployed and experimented on two inbound marketing channels.FindingsResearch examining machine learning implementation and operationalisation within the large consumer base is seemingly silent. This paper bridges this gap by developing and implementing a machine learning adaptive engine across two inbound marketing channels. The performance of the inbound channels revealed the significant importance of digital campaigns that are driven by machine learning algorithms. Machine learning techniques can be well positioned as an integral part of a large consumer base marketing operations with real-time one-to-one marketing capability.Research limitations/implicationsThe study explores the use of machine learning, a cutting-edge subset of artificial intelligence (AI), to drive consumer business revenue across different marketing channels. Further research should explore these marketing channels in greater depth by considering other branches of AI in driving consumer business revenue.Practical implicationsThis study demonstrates the value, ease and application of a machine learning deployment in a consumer business with a large customer base in driving business revenue. It also shows customers' practical response to offerings across channels and the importance of the digital channel to firms with a large customer base.Originality/valueThe paper defines how machine learning extracts can be deployed and operationalised by marketers to drive business revenue. This approach is unique, realistic, easy to deploy and will guide future research in this space.

Implementation of machine learning in engines: evidence from designs, performances and fault diagnosis

Implementation of machine learning in engines: evidence from designs, performances and fault diagnosis

Powder Bed Additive Manufacturing Using Machine Learning Algorithms for Multidisciplinary Applications: A Review and Outlook

Additive manufacturing overcomes the limitations associated with conventional processes, such as fabricating complex parts, material wastage, and a number of sequential operations. Powder-bed additives fall under the category of additive manufacturing process, which, in recent years, has captured the attention of researchers and scientists working in various fields of science and engineering. Production of powder bed additive manufacturing (PBAM) parts with consistent and predictable properties of powders used during the manufacturing process plays an important role in deciding printed parts' reliability in aeronautical, automobile, biomedical, and healthcare applications. In the PBAM process, the most commonly used powders are polymer, metal, and ceramic, which cannot be effectively used without understanding powder particles' physical, mechanical, and chemical properties. Several metallic powders like titanium, steel, copper, aluminum, and nickel, several polymer polyamides (nylon), polylactide, polycarbonate, glass-filled nylon, epoxy resins, etc., and the most commonly used ceramic powders like aluminum oxide (Al2O) and zirconium oxide (ZrO2) can be utilized depending upon the method being adopted during PBAM process. Adoption of some post-processing techniques for powder, such as grain refinement can also be employed to improve the physical or mechanical properties of powders used for the PBAM process. In this paper, the effect of powder parameters, such as particle size, shape, density, and reusing of powder, etc., on printed parts have been reviewed in detail using characterization techniques such as X-ray computed tomography, scanning electron microscopy, and X-ray photoelectron spectroscopy. This helps to understand the effect of particle size, shape, density, virgin and reused powders, etc., used during the PBAM process. This article has reviewed the selection of appropriate process parameters like laser power, scanning speed, hatch spacing, and layer thickness and their effects on various mechanical or physical properties, such as tensile strength, hardness, and the effect of porosity, along with the microstructure evolution. One of the drawbacks of additive manufacturing is the variability in the quality of printed parts, which can be eliminated by monitoring the process using machine learning techniques. Also, the prediction of the best combination of process parameters using some advanced machine learning algorithms (MLA), like random forest, k nearest neighbors, and support vector machine, can be effectively utilized to quantify the performance parameters in the PBAM process. Thus, implementing machine learning in the additive manufacturing process not only helps to learn the fundamentals but helps to identify, predict and help to make actionable recommendations that help optimize printed parts quality. The performance of various MLAs has been evaluated and compared for projecting future research directions and suggestions. In the last part of this article, multidisciplinary applications of the PBAM process have been reviewed in detail. Additive manufacturing processes carried out by using conventional machines, called hybrid additive manufacturing, have also been reviewed by discussing their methods and arrangements in detail.

Fair prediction sets through multi-objective hyperparameter optimization

The widespread implementation of machine learning in safety-critical domains has raised ethical concerns regarding algorithmic discrimination. In such settings, the integration of fairness-aware algorithms with uncertainty quantification tools enables the development of reliable and safe decision-making. In this paper, we introduce a novel methodology that combines conformal prediction, offering rigorous prediction sets, with multi-objective optimization via evolutionary learning. The proposed meta-algorithm optimizes the hyperparameter configuration of classifiers to produce confidence predictors that balance efficiency and equalized coverage guarantees, addressing fairness concerns related to sensitive attributes. We empirically evaluate our methodology with four real-world problems and demonstrate its efficacy in exploring this trade-off and producing a repertoire of Pareto optimal conformal predictors. In this way, our contribution offers different modeling alternatives from which to choose depending on the policy adopted by stakeholders, thus illustrating its capability to enhance equitable decision-making.

Guiding practitioners of road freight transport to implement machine learning for operational planning tasks

Guiding practitioners of road freight transport to implement machine learning for operational planning tasks

Artificial Intelligence in Agriculture: A Technical Analysis of Precision Farming Systems

This technical article comprehensively explores the integration of Artificial Intelligence in agricultural systems, with a primary focus on precision farming methodologies and their practical implementation. It conducts an in-depth examination of the technological framework underpinning agricultural AI systems, encompassing sophisticated data acquisition infrastructure and advanced machine learning implementations. Through detailed analysis, the article investigates the critical technical components of precision farming systems, emphasizing the integration of sensor technologies spanning both aerial and ground-based platforms, alongside robust data analytics architectures. It systematically addresses the prevalent challenges in agricultural data integration and system scalability, presenting innovative solutions for these complex issues. Furthermore, the article explores emerging technological developments in the field, including cutting-edge sensing technologies and AI algorithm enhancements, offering valuable insights into future trajectories of agricultural AI applications. The comprehensive article illuminates how AI technologies are fundamentally transforming traditional farming practices through the implementation of sophisticated automated decision-making processes and enhanced resource management capabilities, ultimately contributing to more efficient and sustainable agricultural operations

New Predictive Models for the Computation of Reinforced Concrete Columns Shear Strength

The assessment methods for estimating the behavior of the complex mechanics of reinforced concrete (RC) structural elements were primarily based on experimental investigation, followed by the collective evaluation of experimental databases from the available literature. There is still a lot of uncertainty in relation to the strength and deformability criteria that have been derived from tests due to the differences in the experimental test setups of the individual research studies that are being fed into the databases used to derive predictive models. This research work focuses on structural elements that exhibit pronounced strength degradation with plastic deformation and brittle failure characteristics. The study’s focus is on evaluating existing models that predict the shear strength of RC columns, which take into account important factors including the structural element’s ductility and axial load, as well as the contributions of specific resistance mechanisms like that of concrete, transverse, and longitudinal reinforcement. Significantly improved predictive models are proposed herein through the implementation of machine learning (ML) algorithms on refined datasets. Three ML models, LREGR, POLYREG-HYT, and XGBoost-HYT-CV, were used to develop different predictive models that were able to compute the shear strength of RC columns. According to the numerical findings, POLYREG-HYT- and XGBoost-HYT-CV-derived models outperformed other ML models in predicting the shear strength of rectangular RC columns with the correlation coefficient having a value R greater than 99% and minimal errors. It was also found that the newly proposed predictive model derived a 2-fold improvement in terms of the correlation coefficient compared to the best available equation in international literature.

Why implementing machine learning algorithms in the clinic is not a plug-and-play solution: a simulation study of a machine learning algorithm for acute leukaemia subtype diagnosis.

Artificial intelligence (AI) and machine learning (ML) algorithms have shown great promise in clinical medicine. Despite the increasing number of published algorithms, most remain unvalidated in real-world clinical settings. This study aims to simulate the practical implementation challenges of a recently developed ML algorithm, AI-PAL, designed for the diagnosis of acute leukaemia and report on its performance. We conducted a detailed simulation of the AI-PAL algorithm's implementation at the University Hospital Essen. Cohort building was performed using our Fast Healthcare Interoperability Resources (FHIR) database, identifying all initially diagnosed patients with acute leukaemia and selected differential diagnoses. The algorithm's performance was assessed by reproducing the original study's results. The AI-PAL algorithm demonstrated significantly lower performance in our simulated clinical implementation compared to prior published results. The area under the receiver operating characteristic curve for acute lymphoblastic leukaemia dropped to 0.67 (95% CI: 0.61-0.73) and for acute myeloid leukaemia to 0.71 (95% CI: 0.65-0.76). The recalibration of probability cutoffs determining confident diagnoses increased the number of confident positive diagnosis for acute leukaemia from 98 to 160, highlighting the necessity of local validation and adjustments. The findings underscore the challenges of implementing ML algorithms in clinical practice. Despite robust development and validation in research settings, ML models like AI-PAL may require significant adjustments and recalibration to maintain performance in different clinical settings. Our results suggest that clinical decision support algorithms should undergo local performance validation before integration into routine care to ensure reliability and safety. This study was supported by the DFG-cofounded UMEA Clinician Scientist Program and the Ministry of Culture and Science of the State of North Rhine-Westphalia.

Risk Assessment and Detection of Fraudulent Claims in Insurance Systems with Machine Learning Approaches

Fraudulent insurance claims pose a significant challenge to the sustainability and efficiency of insurance systems, leading to substantial financial losses and undermining trust between insurers and policyholders. The complexity and volume of contemporary data make it difficult for traditional fraud detection techniques like manual assessments to handle. To detect fraud in insurance claims, this study investigates the use of machine learning approaches, such as Random Forest and Artificial Neural Networks (ANN). Using a structured methodology, models were trained on historical claim data and assessed using accuracy, F1-score, recall, and precision. The Random Forest algorithm's accuracy was 94%, according to the results, while the ANN performed better on controlled datasets. Additionally, feature importance analysis identified key predictors, such as claim amount and submission frequency, providing actionable insights for fraud prevention strategies. The implementation of machine learning into claim management systems offers a scalable, accurate, and cost-effective solution to combating fraud. Future work will focus on testing with larger datasets and exploring hybrid approaches to enhance robustness and adaptability.

Machine Learning Prediction of the Corrosion Rate of Zinc-Based Alloys Containing Copper, Lithium, Magnesium, and Silver.

Implementation of machine learning (ML) techniques in materials science often requires large data sets. However, a proper choice of features and regression methods allows the construction of accurate ML models able to work with a relatively small data set. In this work, an extensive, although still limited, experimental data set of corrosion-related properties of Zn-based alloys used in biomedicine was created. On the basis of this data set, a robust and accurate model was built to predict the corrosion behavior of Zn-based alloys. This work highlights the effectiveness of ML methods for assessing the corrosion behavior of Zn-based alloys, which can facilitate their application in bioimplants.

Effective DDoS attack detection in software-defined vehicular networks using statistical flow analysis and machine learning.

Vehicular Networks (VN) utilizing Software Defined Networking (SDN) have garnered significant attention recently, paralleling the advancements in wireless networks. VN are deployed to optimize traffic flow, enhance the driving experience, and ensure road safety. However, VN are vulnerable to Distributed Denial of Service (DDoS) attacks, posing severe threats in the contemporary Internet landscape. With the surge in Internet traffic, this study proposes novel methodologies for effectively detecting DDoS attacks within Software-Defined Vehicular Networks (SDVN), wherein attackers commandeer compromised nodes to monopolize network resources, disrupting communication among vehicles and between vehicles and infrastructure. The proposed methodology aims to: (i) analyze statistical flow and compute entropy, and (ii) implement Machine Learning (ML) algorithms within SDN Intrusion Detection Systems for Internet of Things (IoT) environments. Additionally, the approach distinguishes between reconnaissance, Denial of Service (DoS), and DDoS traffic by addressing the challenges of imbalanced and overfitting dataset traces. One of the significant challenges in this integration is managing the computational load and ensuring real-time performance. The ML models, especially complex ones like Random Forest, require substantial processing power, which necessitates efficient data handling and possibly leveraging edge computing resources to reduce latency. Ensuring scalability and maintaining high detection accuracy as network traffic grows and evolves is another critical challenge. By leveraging a minimal subset of features from a given dataset, a comparative study is conducted to determine the optimal sample size for maximizing model accuracy. Further, the study evaluates the impact of various dataset attributes on performance thresholds. The K-nearest Neighbor, Random Forest, and Logistic Regression supervised ML classifiers are assessed using the BoT-IoT dataset. The results indicate that the Random Forest classifier achieves superior performance metrics, with Precision, F1-score, Accuracy, and Recall rates of 92%, 92%, 91%, and 90%, respectively, over five iterations.

Application of Machine Learning Techniques in Calibration and Data Reduction of Multihole Probes

Abstract This work presents procedures for implementing machine learning methods into existing algorithms for multihole probe calibration and data reduction. It demonstrates that using artificial neural networks (ANNs) can decrease the amount of calibration data needed to achieve a specific calibration uncertainty by over 50%, while also significantly reducing data reduction times. Instead of surface fitting methods, ANNs are employed. Initially, directional calibration coefficients related to flow angles are computed based on pressure measurements, and then these flow angles serve as input parameters for subsequent ANNs to iteratively define Mach number, static pressure, and total pressure. In an alternative approach, new calibration coefficients directly relate pressure measurements from the five-hole probe to the quantities of interest, thereby eliminating the need for iterative algorithms used in conventional surface fitting methods. This method offers several advantages: an average increase of less than 1% in calibration uncertainty for flow angles and a significant reduction in data reduction times to a few seconds on average. Additionally, the methodology is confirmed to avoid both overfitting and underfitting.

Numerical Homogenization of Orthotropic Functionally Graded Periodic Cellular Materials: Method Development and Implementation.

This study advances the state of the art by computing the macroscopic elastic properties of 2D periodic functionally graded microcellular materials, incorporating both isotropic and orthotropic solid phases, as seen in additively manufactured components. This is achieved through numerical homogenization and several novel MATLAB implementations (known in this study as Cellular_Solid, Homogenize_test, homogenize_ortho, and Homogenize_test_ortho_principal). The developed codes in the current work treat each cell as a material point, compute the corresponding cell elasticity tensor using numerical homogenization, and assign it to that specific point. This is conducted based on the principle of scale separation, which is a fundamental concept in homogenization theory. Then, by deriving a fit function that maps the entire material domain, the homogenized material properties are predicted at any desired point. It is shown that this method is very capable of capturing the effects of orthotropy during the solid phase of the material and that it effectively accounts for the influence of void geometry on the macroscopic anisotropies, since the obtained elasticity tensor has different E1 and E2 values. Also, it is revealed that the complexity of the void patterns and the intensity of the void size changes from one cell to another can significantly affect the overall error in terms of the predicted material properties. As the stochasticity in the void sizes increases, the error also tends to increase, since it becomes more challenging to interpolate the data accurately. Therefore, utilizing advanced computational techniques, such as more sophisticated fitting methods like the Fourier series, and implementing machine learning algorithms can significantly improve the overall accuracy of the results. Furthermore, the developed codes can easily be extended to accommodate the homogenization of composite materials incorporating multiple orthotropic phases. This implementation is limited to periodic void distributions and currently supports circular, rectangular, square, and hexagonal void shapes.

Digital Assistance Systems to Implement Machine Learning in Manufacturing: A Systematic Review

Digital Assistance Systems to Implement Machine Learning in Manufacturing: A Systematic Review

The machine learning implementation and theoretical foundations of intelligent control of accelerator particle beams

The machine learning implementation and theoretical foundations of intelligent control of accelerator particle beams

Predictive Models for Interest Rate Forecasting Using Machine Learning: A Comparative Analysis and Practical Application

Forecasting interest rates is a fundamental task in financial planning, investment strategies, and policy-making. Traditional statistical models, while widely used, often fail to adequately capture complex non-linear relationships and temporal dependencies inherent in financial data. This study addresses these limitations by exploring the potential of machine learning models to improve the accuracy and reliability of interest rate forecasts. The primary objective of this research is to evaluate and compare the performance of multiple machine learning models, including linear regression, support vector machines, and deep learning techniques, in predicting interest rate trends. Historical data spanning two decades was collected and preprocessed, ensuring data quality and consistency. The models were trained and tested on this dataset using well-defined evaluation metrics such as mean absolute error and root mean squared error to ensure robust performance assessments. The results revealed that machine learning approaches, particularly deep learning models, outperformed traditional methods in capturing complex patterns and delivering more accurate forecasts. The findings further discuss the practical implications of implementing machine learning techniques in real-world financial contexts, highlighting both opportunities and challenges. In conclusion, this study provides actionable insights and a robust framework for integrating machine learning into interest rate forecasting, contributing to the advancement of predictive modeling in finance.

Predictive preoperative factors of anastomotic fistula after left colon cancer surgery: implementation of machine learning

Predictive preoperative factors of anastomotic fistula after left colon cancer surgery: implementation of machine learning

EEG-Based User Identification using Machine Learning and Deep Learning Approaches

Abstract In recent years, Electroencephalogram (EEG) based user authentication systems have gained significant interest as an innovative approach for identity verification. EEGs are considered to be a novel biometric attribute due to the individuality of each person’s cerebral activity patterns. This work explores the feasibility and efficiency of utilizing EEG signals, generated in response to emotional stimuli, for user authentication applications, by implementing Machine Learning (ML) and Deep Learning (DL) approaches. Support Vector Machine (SVM), Random Forest (RF) classifier and 1D Convolution Neural Network (CNN) were employed to evaluate and compare the performance of EEG-based user authentication for two publicly available EEG datasets, namely DEAP and DENS database. The performance of EEGbased user authentication was significantly high in LAHV emotional state for DENS dataset, achieving an accuracy of 99.2 % and 92.59 % with SVM and modified 1D CNN, respectively.

Predicting adolescent psychopathology from early life factors: A machine learning tutorial

ObjectiveThe successful implementation and interpretation of machine learning (ML) models in epidemiological studies can be challenging without an extensive programming background. We provide a didactic example of machine learning for risk prediction in this study by determining whether early life factors could be useful for predicting adolescent psychopathology. MethodsIn total, 9643 adolescents ages 9–10 from the Adolescent Brain and Cognitive Development (ABCD) Study were included in ML analysis to predict high Child Behavior Checklist (CBCL) scores (i.e., t-scores ≥ 60). ML models were constructed using a series of predictor combinations (prenatal, family history, sociodemographic) across 5 different algorithms. We assessed ML performance through sensitivity, specificity, F1-score, and area under the curve (AUC) metrics. ResultsA total of 1267 adolescents (13.1 %) were found to have high CBCL scores. The best performing algorithms were elastic net and gradient boosted trees. The best performing elastic net models included prenatal and family history factors (Sensitivity 0.654, Specificity 0.713; AUC 0.742, F1-score 0.401) and prenatal, family, history, and sociodemographic factors (Sensitivity 0.668, Specificity 0.704; AUC 0.745, F1-score 0.402). Across all 5 ML algorithms, family history factors (e.g., either parent had nervous breakdowns, trouble holding jobs/fights/police encounters, and counseling for mental issues) and sociodemographic covariates (e.g., maternal age, child's sex, caregiver income and caregiver education) tended to be better predictors of adolescent psychopathology. The most important prenatal predictors were unplanned pregnancy, birth complications, and pregnancy complications. ConclusionOur results suggest that inclusion of prenatal, family history, and sociodemographic factors in ML models can generate moderately accurate predictions of adolescent psychopathology. Issues associated with model overfitting, hyperparameter tuning, and system seed setting should be considered throughout model training, testing, and validation. Future early risk predictions models may improve with the inclusion of additional relevant covariates.

Self-supervised Pre-training via Multi-view Graph Information Bottleneck for Molecular Property Prediction.

Molecular representation learning has remarkably accelerated the development of drug analysis and discovery. It implements machine learning methods to encode molecule embeddings for diverse downstream drug-related tasks. Due to the scarcity of labeled molecular data, self-supervised molecular pre-training is promising as it can handle large-scale unlabeled molecular data to prompt representation learning. Although many universal graph pre-training methods have been successfully introduced into molecular learning, there still exist some limitations. Many graph augmentation methods, such as atom deletion and bond perturbation, tend to destroy the intrinsic properties and connections of molecules. In addition, identifying subgraphs that are important to specific chemical properties is also challenging for molecular learning. To address these limitations, we propose the self-supervised Molecular Graph Information Bottleneck (MGIB) model for molecular pre-training. MGIB observes molecular graphs from the atom view and the motif view, deploys a learnable graph compression process to extract the core subgraphs, and extends the graph information bottleneck into the self-supervised molecular pre-training framework. Model analysis validates the contribution of the self-supervised graph information bottleneck and illustrates the interpretability of MGIB through the extracted subgraphs. Extensive experiments involving molecular property prediction, including 7 binary classification tasks and 6 regression tasks demonstrate the effectiveness and superiority of our proposed MGIB.