Machine Learning Algorithm Model Research Articles

Experimental investigation and machine learning models for predicting flexural and tensile strengths of recycled concrete: Bridging the gap between sustainable materials and structural design

The sustainability of structural components requires recycled aggregate concrete (RC) to achieve adequate flexural and split tensile strengths for practical use. These strengths depend on mix design and the properties of recycled coarse aggregates (R-CA). The variability in R-CA sources and the inherent heterogeneity of RC complicate strength predictions. No existing model accounts for this variability, leaving a gap between sustainable materials and structural design. This study develops machine learning-based models to predict the flexural and split tensile strengths of RC, regardless of R-CA source and properties. Key factors such as water absorption, effective water-to-cement ratio, coarse aggregate-to-cement ratio, and R-CA replacement ratio are used for predictions. The impact of different R-CA types on RC and natural aggregate concrete (NC) is also experimentally analyzed. A dataset of 353 test results from this study and 33 prior studies is used, and various machine learning algorithms (MLA) are evaluated. Results show a 41 % and 23 % reduction in flexural and split tensile strengths of RC compared to NC, but acid and mechanically treated R-CA can recover up to 94 % and 93 % of NC's strengths, respectively. Among all MLA models, the gradient boost model depicted the highest accuracy in predictions for the flexural and tensile strengths of both RC and NC. This research introduces new equations and a C/C++ tool for predicting RC and NC strengths, contributing to sustainable concrete design and bridging the gap between research and practical application.

Machine Learning Approaches for Predicting Power Conversion Efficiency in Organic Solar Cells: A Comprehensive Review

Organic solar cells (OSCs), renowned for their lightweight, cost efficiency, and adaptability nature, stand out as a promising option for developing renewable energy. Improving the power conversion efficiency (PCE) of OSCs is essential, and researchers are delving into novel materials to achieve this. Traditional approaches are often laborious and costly, highlighting the need for predictive modeling. Machine learning (ML), especially via quantitative structure–property relationship (QSPR) models, is streamlining material development, with a goal to exceed a 20% PCE. In this review, the application of ML in OSCs is explored, and recent studies utilizing ML approaches for PCE prediction are reviewed, encompassing empirical functions, ML algorithms, self‐devised ML frameworks, and the combination with automated experimental technologies. First, the benefits of ML in predicting PCE for OSCs are addressed. Second, the development of high‐efficiency predictive models for both fullerene and nonfullerene acceptors is delved into. The impact of various ML algorithm models on PCE prediction is then assessed, taking into account the construction of predictive models. Moreover, the quality of databases and the selection of descriptors are considered. Databases and descriptors based on experimental studies are further categorized. Finally, prospects for the future development of OSCs are proposed.

Construction and verification of a machine learning-based prediction model of deep vein thrombosis formation after spinal surgery

BackgroundDeep vein thromboembolism (DVT) is a common postoperative complication with high morbidity and mortality rates. However, the safety and effectiveness of using prophylactic anticoagulants for preventing DVT after spinal surgery remain controversial. Hence, it is crucial to predict whether DVT occurs in advance following spinal surgery. The present study aimed to establish a machine learning (ML)-based prediction model of DVT formation following spinal surgery. MethodsWe reviewed the medical records of patients who underwent elective spinal surgery at the Third Affiliated Hospital of Zunyi Medical University (TAHZMU) from January 2020 to December 2022. We ultimately selected the clinical data of 500 patients who met the criteria for elective spinal surgery. The Boruta-SHAP algorithm was used for feature selection, and the SMOTE algorithm was used for data balance. The related risk factors for DVT after spinal surgery were screened and analyzed. Five ML algorithm models were established. The data of 150 patients treated at the Affiliated Hospital of Zunyi Medical University (AHZMU) from July 2023 to October 2023 were used for external verification of the model. The area under the curve (AUC), geometric mean (G-mean), sensitivity, accuracy, specificity, and F1 score were used to evaluate the performance of the models. ResultsThe results revealed that activated partial thromboplastin time (APTT), age, body mass index (BMI), preoperative serum creatinine (Crea), anesthesia time, rocuronium dose, and propofol dose were the seven important characteristic variables for predicting DVT after spinal surgery. Among the five ML models established in this study, the random forest classifier (RF) showed superior performance to the other models in the internal validation set. ConclusionSeven preoperative and intraoperative variables were included in our study to develop an ML-based predictive model for DVT formation following spinal surgery, and this model can be used to assist in clinical evaluation and decision-making.

Evaluation of multiple machine learning models for ASR expansion of concrete

Machine learning algorithms (MLA) show promise in predicting alkali-silica reaction (ASR) expansion. Nevertheless, the predictive performance of different machine learning algorithms (MLAs) varies across various engineering scenarios and datasets. Given that, this study was devoted to constructing an appropriate predictive model for ASR expansion of concrete. Multiple MLAs including evolutionary-based, support vector-based, assembly-based, and neural network-based algorithms, were employed to establish predictive models. An optimization technique, the particle swarm optimization algorithm, was also utilized and integrated with the aforementioned MLA models to effectively search for optimal hyperparameters and topological structures. The predictive performance was evaluated using statistical methods, which showed that the extreme gradient boosting (XGBoost) model had the best fitting goodness among the standard algorithms, with an R2 value of 0.9695. Moreover, the Particle Swarm Optimization (PSO) model significantly improved the predictive performance of all Machine Learning Algorithm (MLA) models. The error distribution analysis was then performed. There are fewer large errors in the predicted values of the PSO-XGBoost model. Monte Carlo simulation and probability distribution functions were used to thoroughly analyze the stability of each model after multiple independent runs. The results show that all models have excellent stability. The PSO-XGBoost model performed the best and was subsequently employed to conduct feature analysis to determine the significance of each input parameter and capture the underlying evidence of ASR expansion.

Machine Learning for Predicting Distant Metastasis of Medullary Thyroid Carcinoma Using the SEER Database.

We aimed to establish an effective machine learning (ML) model for predicting the risk of distant metastasis (DM) in medullary thyroid carcinoma (MTC). Demographic data of MTC patients were extracted from the Surveillance, Epidemiology, and End Results (SEER) database of the National Institutes of Health between 2004 and 2015 to develop six ML algorithm models. Models were evaluated based on accuracy, precision, recall rate, F1-score, and area under the receiver operating characteristic curve (AUC). The association between clinicopathological characteristics and target variables was interpreted. Analyses were performed using traditional logistic regression (LR). In total, 2049 patients were included and 138 developed DM. Multivariable LR showed that age, sex, tumor size, extrathyroidal extension, and lymph node metastasis were predictive features for DM in MTC. Among the six ML models, the random forest (RF) had the best predictability in assessing the risk of DM in MTC, with an accuracy, precision, recall rate, F1-score, and AUC higher than those of the traditional binary LR model. RF was superior to traditional LR in predicting the risk of DM in MTC and can provide a valuable reference for clinicians in decision-making.

Predicting dengue transmission rates by comparing different machine learning models with vector indices and meteorological data

Machine learning algorithms (ML) are receiving a lot of attention in the development of predictive models for monitoring dengue transmission rates. Previous work has focused only on specific weather variables and algorithms, and there is still a need for a model that uses more variables and algorithms that have higher performance. In this study, we use vector indices and meteorological data as predictors to develop the ML models. We trained and validated seven ML algorithms, including an ensemble ML method, and compared their performance using the receiver operating characteristic (ROC) with the area under the curve (AUC), accuracy and F1 score. Our results show that an ensemble ML such as XG Boost, AdaBoost and Random Forest perform better than the logistics regression, Naïve Bayens, decision tree, and support vector machine (SVM), with XGBoost having the highest AUC, accuracy and F1 score. Analysis of the importance of the variables showed that the container index was the least important. By removing this variable, the ML models improved their performance by at least 6% in AUC and F1 score. Our result provides a framework for future studies on the use of predictive models in the development of an early warning system.

Intrusion detection system for detecting distributed denial of service attacks using machine learning algorithms

<span>Today, the creation of more effective intrusion detection system (IDS) has become crucial due to the rise in computer malware. Ensuring the availability of the system is an important component of information security and the most important requirement of any network. Recently the machine learning algorithm (ML) has been used to improve intrusion detection over the network. It is currently necessary to release an updated version of these systems. The presented work aimed to build a reliable and accurate IDS based on ML to classify and prevent distributed denial of service attacks to protect any system working on the network from temporary or complete system failure. We presented five ML models to create the proposed distributed denial-of-services attack (DDoS)-IDS, including (decision tree, random forest, logistic regression, support vector machine, and multi-layer neural network) which were trained and evaluated using the CIC-IDS-2018 dataset. Furthermore, principal component analysis (PCA) was used to reduce the dimensionality of the dataset. According to the classification results, the proposed multi-layer neural network model reached optimal performance for detecting DDoS attacks and achieved classification accuracy at 99.9992%.</span>

Reliability assessment of open-source multiscale landslide susceptibility maps and effects of their fusion

ABSTRACT Several landslide susceptibility (LS) maps at various scales of analysis have been performed with specific zoning purposes and techniques. Supervised machine learning algorithms (ML) have become one of the most diffused techniques for landslide prediction, whose reliability is firmly based on the quality of input data. Site-specific landslide inventories are often more accurate and complete than national or worldwide databases. For these reasons, detailed landslide inventory and predisposing variables must be collected to derive reliable LS products. However, high-quality data are often rare, and risk managers must consider lower-resolution available products with no more than informative purposes. In this work, we compared different ML models to select the most accurate for large-scale LS assessment within the Municipality of Rome. The ExtraTreesClassifier outperformed the others reaching an average F1-score of 0.896. Thereafter, we addressed the reliability of open-source LS maps at different scales of analysis (global to regional) by means of statistical and spatial analysis. The obtained results shed light on the difference in hazard zoning depending on the scale and mapping unit. An approach for low-resolution LS data fusion was attempted, assessing the importance of the adopted criteria, which increased the ability to detect occurred landslides while maintaining precision.

ARTIFICIAL INTELLIGENCE BASED RATING OF CARPAL TUNNEL SYNDROME EFFICACY IN CLINICAL DIAGNOSIS

Objective: The most common entrapment neuropathy seen by the clinician is Carpal tunnel syndrome (CTS). CTS is graded as mild, moderate, and severe according to the results obtained on electroneuromyography (ENMG) by clinicians. We aimed to show the effectiveness of the use of artificial intelligence in clinical diagnosis in the grading of CTS.
 Methods: In our study, the data of 315 people with a pre-diagnosis of CTS were used and classified into four classes based on AI as CTS grade. Machine Learning (ML) algorithms Ensemble, Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Decision Tree (Tree) algorithms were used in classification processes. 10% Hold-out validation was used and the learning rate was determined as 0.1. As a result of the classification, accuracy, precision, sensitivity, specificity, and F1-score performance values were obtained. 
 Results: SVM made the best estimation and KNN made the worst estimation in the 0 class. The best estimate in class 1 belongs to the Support Vector Machine. Ensemble and Tree made the best guesses in the 2nd and 3rd grades. In our study, the best algorithm with an overall success rate is SVM with 93.55%. 
 Conclusions: The results showed that ML algorithm models consistently provided better predictive results and would assist physicians in determining the medical treatment modality of CTS. Artificial intelligence (AI) techniques are reliable methods that assist clinicians to deliver quality healthcare.

Gender Estimation with Parameters Obtained From the Upper Dental Arcade by Using Machine Learning Algorithms and Artificial Neural Networks

Objective: The aim of this study is to estimate gender with parameters obtained from the upper dental arcade by using machine learning algorithms and artificial neural networks. Methods: The study was conducted on cone-beamed computed tomography images of 176 individuals between the ages of 18 and 55 who did not have any pathologies or surgical interventions in their upper dental arcade. The images obtained were transferred to RadiAnt DICOM Viewer program in Digital Imaging and Communications in Medicine format and all images were brought to orthogonal plane by applying 3D Curved Multiplanar Reconstruction. Length and curvature length measurements were performed on these images brought to orthogonal plane. The data obtained were used in machine learning algorithms (ML) and artificial neural networks input and rates of gender estimation were shown. Results: In the study, an accuracy ratio of 0.86 was found with ML models linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), logistic regression (LR) algorithm and an accuracy ratio of 0.86 was found with random forest (RF) algorithm. It was found with SHAP analyser of RF algorithm that the parameter of width at the level of 3rd molar teeth contributed the most to gender. An accuracy rate of 0.92 was found as a result of training for 500 times with multilayer perceptron classifier (MLCP), which is an artificial neural network (ANN) model. Conclusion: As a result of our study, it was found that the parameters obtained from the upper dental arcade showed a high accuracy in estimation of gender. In this context, we believe that the present study will make important contributions to forensic sciences.

517-P: Machine Learning Algorithms Using Electronic Medical Record and Ultrasonography of Lower Extremities for 10-Year Risk Assessment of Diabetic Foot Ulcer in China

Objective: Accurate prediction of diabetic foot ulcer (DFU) occurrence in patients with type 2 diabetes mellitus (T2DM) is crucial for determining appropriate treatment strategy and avoiding amputation and infection. Here, we explored machine learning algorithms (ML) of electronic medical record (EMR) for predicting the incident of DFU. Methods: In this retrospective longitudinal observational cohort, 26579 cases from T2DM database of The Third Affiliated Hospital of Sun Yatsen University from Jan 2011 to Dec 2019 were screened and finally 2770 patients without DFU were enrolled. During 5.6 years of follow-up, 178 patients occurred DFU. 44 clinical parameters including EMR and imaging finding were used to develop a DFU risk prediction model using XGBoost, RF, and LR. Results: ML models were established in 178 patients with DFU (110 male, mean 61.1±12.4 years), and 2591 patients without DFU (1424 male, mean 59.0±12.1 years). XGBoost model achieved the best performance in predicting DFU occurrence with an AUC of 0.94, which is better than the performance of RF with AUC of 0.91and LR with AUC of 0.89. A DFU incident significantly increased in high-risk patients (>10%) in the next 10 years. Conclusion: Our study provided a simple and effective risk tool for the prediction of 10-year DFU risk, which can help professionals clinical screening and decision-making for T2DM patients. Disclosure L.Wu: None. Y.Liang: None. X.He: None. Y.Chen: None.

Impact of Machine Learning and Prediction Models in the Diagnosis of Oral Health Conditions

Introduction: Recent developments in data science and the employment of machine learning algorithms (ML) have revolutionized health sciences in the prediction of diseases using laboratory data. Oral diseases are observed in all age groups and are estimated to affect about a 3.5billion people as per WHO 2022 statistics. Using the existing diagnostic data and taking advantage of ML and prediction models would benefit developing a prediction model for diagnosing oral diseases. Hence, it is quite essential to understand the basic terminologies used in the prediction model.
 Methods: We retrieve various research papers using Scopus, PubMed, and google scholar databases, where prediction models were used in dentistry. The idea of this review is to explore current models, model validation, discrimination, calibration, and bootstrapping methods used in prediction models for oral diseases.
 Results: The current advancement of ML techniques plays a significant task in the diagnosis and prognosis of oral diseases.
 Conclusion: The use of prediction models using ML techniques can improve the accuracy of the treatment methods in oral health. This article aims to provide the required framework, data sets, and methodology to build ML and prediction models for oral diseases.

Lifetime prediction of electronic devices based on the P-stacking machine learning model

Nowadays the data-driven artificial intelligence (AI) and machine learning (ML) provide novel approaches for the effective lifetime prediction of the electronic devices with complicated mechanisms and multiple controlling factors. However, the existing ML prediction models cannot process both high accuracy and high efficiency at the meantime. To address this problem, this paper proposes the new P-Stacking ML algorithm model, which combines the Pearson correlation analysis and the Stacking multi-model fusion. The Pearson correlation analysis is first applied to select the multiple base learner models with weak correlations for the following Stacking multi-model fusion. Two neutral network layers are built to perform the Stacking algorithm. In this work, two distinct types of electronic devices are utilized to verify the effectiveness of the P-Stacking ML model, which are the IGBT devices and the lithium-ion batteries respectively. For the IGBT lifetime prediction, the results have shown that compared with the long-term and short-term memory neural network (LSTM) model in the previous literature, the mean square error (MSE) of the P-Stacking ML model is improved by 6 %, and the average model training time is reduced by 90 %. Moreover, for the lithium batteries, the lifetime prediction accuracy of the P-Stacking model is increased by 65 %, and the training time is decreased by 90 %. The results demonstrated that the P-Stacking ML model can significantly improve both the prediction accuracy and efficiency simultaneously.

GeoZ: a Region-Based Visualization of Clustering Algorithms

The spatial display of clustered data using machine learning (ML) as regions (bordered areas) is currently unfeasible. This problem is commonly encountered in various research fields that utilize clustering algorithms in their workflow. We present in this study an approach utilizing ML algorithm models that can be trained to any specific dataset to produce decision boundaries. These boundaries are overlaid onto the geographic coordinate system (GCS) to generate geographic clustering regions. The proposed approach is implemented in the Python Package Index (PyPI) as a geovisualization library called geographic decision zones (GeoZ). The efficiency of GeoZ was tested using a dataset of groundwater wells in the State of California. We experimented with 13 different ML models to determine the best model that predicts the existing regional distribution (subbasins). The support vector machine (SVM) algorithm produced a relatively high accuracy score and fulfilled the required criteria better than the other models. Consequently, the tested SVM model with optimized parameters was implemented in the GeoZ open-source library. However, it is important to note that limitations in the application of GeoZ may arise from the nature of the SVM algorithm, as well as the volume, discontinuity, and distribution of the data. We have attempted to address these limitations through various suggestions and solutions.

An Improved Machine Learning-Based Approach to Assess the Microbial Diversity in Major North Indian River Ecosystems.

The rapidly evolving high-throughput sequencing (HTS) technologies generate voluminous genomic and metagenomic sequences, which can help classify the microbial communities with high accuracy in many ecosystems. Conventionally, the rule-based binning techniques are used to classify the contigs or scaffolds based on either sequence composition or sequence similarity. However, the accurate classification of the microbial communities remains a major challenge due to massive data volumes at hand as well as a requirement of efficient binning methods and classification algorithms. Therefore, we attempted here to implement iterative K-Means clustering for the initial binning of metagenomics sequences and applied various machine learning algorithms (MLAs) to classify the newly identified unknown microbes. The cluster annotation was achieved through the BLAST program of NCBI, which resulted in the grouping of assembled scaffolds into five classes, i.e., bacteria, archaea, eukaryota, viruses and others. The annotated cluster sequences were used to train machine learning algorithms (MLAs) to develop prediction models to classify unknown metagenomic sequences. In this study, we used metagenomic datasets of samples collected from the Ganga (Kanpur and Farakka) and the Yamuna (Delhi) rivers in India for clustering and training the MLA models. Further, the performance of MLAs was evaluated by 10-fold cross validation. The results revealed that the developed model based on the Random Forest had a superior performance compared to the other considered learning algorithms. The proposed method can be used for annotating the metagenomic scaffolds/contigs being complementary to existing methods of metagenomic data analysis. An offline predictor source code with the best prediction model is available at (https://github.com/Nalinikanta7/metagenomics).

Machine learning for forecasting a photovoltaic (PV) generation system

To mitigate the carbon print of buildings, they should have on-site renewable energy generation systems to supply energy for the buildings without relying on the national grid. Renewable generation sources rely on weather conditions and are therefore difficult to rely on as the only source of energy. Photovoltaic (PV) is forecasted through machine learning algorithms (MLA), but different methods have varied accuracy and have different training requirements such as more inputs or more data in general. No previous research has concluded an optimal MLA but to better apply them to PV systems, this must be established. To conclude an optimal MLA for a particular application, the dataset and required outputs must be determined, and how they affect the performance of the algorithm must be evaluated. The aim of this work is to compare benchmark MLA's through accuracy and usability for an operational University campus located in central Manchester, in the north of England. The MLA's including random forest (RF), neural networks (NN), support vector machines (SVM), and linear regression (LR) have been employed to forecast the PV system. If the power output of the renewables is accurately forecasted, a building management system (BMS) can be equipped to optimise on-site renewable energy generation. To accomplish this, sixty-four MLA models are created in total for forecasting at multiple horizons and dataset sizes which are validated against real-time data. Results in this work revealed that the RF algorithms have the lowest average error of the multiple tests at 32 root mean squared error (RMSE), whereas SVM, LR, and NN showed at 32.3 RMSE, 36.5 RMSE, and 38.9 RMSE respectively. Errors between forecasted and actual results are recorded in RMSE whereas changes in error are shown in mean actual percentage error (MAPE) to show the changes with respect to the original value. No MLA outperforms all others for accuracy and for requiring less data. No previous research is conducted to evaluate the performance of various MLA PV forecasting models through various sized data sets with critical analysis on the results. The comparison of benchmark algorithms when forecasting the PV generation of a local system allows the critical analysis of the models' accuracy and surrounding characteristics.

Shear Strength Prediction of Slender Concrete Beams Reinforced with FRP Rebar Using Data-Driven Machine Learning Algorithms

Estimating the shear strength of a fiber-reinforced polymer (FRP)–reinforced-concrete (RC) beam is a complex task that depends on multiple design variables. The use of FRP bars has emerged as a promising alternative to diminish the corrosion problems that are associated with steel reinforcement in adverse environments; however, an accurate and reliable method of shear strength prediction is needed to ensure the economical use of materials and robust designs. Several optimized design equations are available in the literature; however, when utilizing these equations a substantial difference is observed between the predicted outcome (Vpred) and the experimental shear strength (Vexp) result. Therefore, this paper presented a novel approach toward implementing machine learning (ML) algorithms to accurately estimate the shear strength of FRP–RC beams. A large database that consisted of 302 shear test results on FRP-reinforced slender concrete beams without stirrup was collected from the literature to formulate the most efficient prediction model. The performance of each ML algorithm model was compared with the existing design provisions and models. The model interpretation was performed through feature importance analysis to explain the model output compared with a black box. The proposed data-driven ML models demonstrated a high level of accuracy and excellent performance and were superior to the existing shear strength models. In addition, a simple graphical user interface (GUI) was developed to aid practicing engineers when estimating shear strength without the need for complicated design procedures.

Application of logistic regression and machine learning methods for idiopathic inflammatory myopathies malignancy prediction.

Malignancy is related to idiopathic inflammatory myopathies (IIM) and leads to a poor prognosis. Early prediction of malignancy is thought to improve the prognosis. However, predictive models have rarely been reported in IIM. Herein, we aimed to establish and use a machine learning (ML) algorithm to predict the possible risk factors for malignancy in IIM patients. We retrospectively reviewed the medical records of 168 patients diagnosed with IIM in Shantou Central hospital, from 2013 to 2021. We randomly divided patients into two groups, the training sets (70%) for construction of the prediction model, and the validation sets (30%) for evaluation of model performance. We constructed six types of ML algorithms models and the AUC of ROC curves were used to describe the efficacy of the model. Finally, we set up a web version using the best prediction model to make it more generally available. According to the multi-variable regression analysis, three predictors were found to be the risk factors to establish the prediction model, including age, ALT<80U/L, and anti-TIF1-γ, and ILD was found to be a protective factor. Compared with five other ML algorithms models, the traditional algorithm logistic regression (LR) model was as good or better than the other models to predict malignancy in IIM. The AUC of the ROC using LR was 0.900 in the training set and 0.784 in the validation set. We selected the LR model as the final prediction model. Accordingly, a nomogram was constructed using the above four factors. A web version was built and can be visited on the website or acquired by scanning the QR code. The LR algorithm appears to be a good predictor of malignancy and may help clinicians screen, evaluate and follow up high-risk patients with IIM.

Applications of Machine Learning Algorithms for Photovoltaic Fault Detection: a Review

Over the years, the boom of technology has caused the accumulation of a large amount of data, famously known as big data, in every field of life. Traditional methods have failed to analyse such a huge pile of data due to outdated techniques. In recent times, the use of photovoltaic systems has risen worldwide. The arena Photovoltaic (PV) system has witnessed the same unprecedented expansion of data owing to the associated monitoring systems. However, the faults created within the PV system cannot be detected, classified, or predicted by using conventional techniques. This necessitates the use of modern techniques such as Machine Learning. Its powerful algorithms, such as artificial neural networks (ANN), help in the accurate detection and classification of faults in the PV system. This review paper introduces and evaluates the applications of Machine Learning (ML) algorithms in PV fault detection. It provides a brief overview of Machine Learning and its concepts along with various widely used ML algorithms. This review various peer-reviewed studies to investigate various models of ML algorithms in the PV system with the main focus on its fault detection accuracy and efficiency.

Machine learning–based mortality prediction models using national liver transplantation registries are feasible but have limited utility across countries

Many countries curate national registries of liver transplant (LT) data. These registries are often used to generate predictive models; however, potential performance and transferability of these models remain unclear. We used data from 3 national registries and developed machine learning algorithm (MLA)–based models to predict 90-day post-LT mortality within and across countries. Predictive performance and external validity of each model were assessed. Prospectively collected data of adult patients (aged ≥18 years) who underwent primary LTs between January 2008 and December 2018 from the Canadian Organ Replacement Registry (Canada), National Health Service Blood and Transplantation (United Kingdom), and United Network for Organ Sharing (United States) were used to develop MLA models to predict 90-day post-LT mortality. Models were developed using each registry individually (based on variables inherent to the individual databases) and using all 3 registries combined (variables in common between the registries [harmonized]). The model performance was evaluated using area under the receiver operating characteristic (AUROC) curve. The number of patients included was as follows: Canada, n = 1214; the United Kingdom, n = 5287; and the United States, n = 59,558. The best performing MLA-based model was ridge regression across both individual registries and harmonized data sets. Model performance diminished from individualized to the harmonized registries, especially in Canada (individualized ridge: AUROC, 0.74; range, 0.73-0.74; harmonized: AUROC, 0.68; range, 0.50-0.73) and US (individualized ridge: AUROC, 0.71; range, 0.70-0.71; harmonized: AUROC, 0.66; range, 0.66-0.66) data sets. External model performance across countries was poor overall. MLA-based models yield a fair discriminatory potential when used within individual databases. However, the external validity of these models is poor when applied across countries. Standardization of registry-based variables could facilitate the added value of MLA-based models in informing decision making in future LTs.