Comparison Of Models Research Articles

Multilevel Latent Differential Structural Equation Model with Short Time Series and Time-Varying Covariates: A Comparison of Frequentist and Bayesian Estimators

Continuous-time modeling using differential equations is a promising technique to model change processes with longitudinal data. Among ways to fit this model, the Latent Differential Structural Equation Modeling (LDSEM) approach defines latent derivative variables within a structural equation modeling (SEM) framework, thereby allowing researchers to leverage advantages of the SEM framework for model building, estimation, inference, and comparison purposes. Still, a few issues remain unresolved, including performance of multilevel variations of the LDSEM under short time lengths (e.g., 14 time points), particularly when coupled multivariate processes and time-varying covariates are involved. Additionally, the possibility of using Bayesian estimation to facilitate the estimation of multilevel LDSEM (M-LDSEM) models with complex and higher-dimensional random effect structures has not been investigated. We present a series of Monte Carlo simulations to evaluate three possible approaches to fitting M-LDSEM, including: frequentist single-level and two-level robust estimators and Bayesian two-level estimator. Our findings suggested that the Bayesian approach outperformed other frequentist approaches. The effects of time-varying covariates are well recovered, and coupling parameters are the least biased especially using higher-order derivative information with the Bayesian estimator. Finally, an empirical example is provided to show the applicability of the approach.

Small-signal modeling comparison of dual active bridge converter

The dual active bridge (DAB) converter is a versatile DC-DC converter, which has AC components in the high-frequency transformer within the converter topology as well as DC components at the input and output sides. These AC components are constructed from infinite sinusoids of multiples of the converter's switching frequency. Different approaches have been proposed to model the DAB converter. The generalized averaging modeling (GAM) includes only the first order harmonic of the inductor current with the DC components of the input and output voltages and currents. For better model accuracy, harmonic-domain models have been employed, where the model includes all the harmonics of all the model state variables up to an arbitrary harmonic index. While the harmonic model provides a systematic approach to include the harmonics, this increases the model complexity. To avoid this complexity, the flexible harmonic model is used where harmonics can be selected per each state variable without needing to be in successive order. The aforementioned modeling approaches have been implemented on the DAB converter. This paper aims to compare between the dynamic performance of GAM and the harmonic modeling approaches under different disturbances, where the GAM approach provided a simple model with acceptable results that can be used for controller design, while the LTP model provided a more complex model with better accuracy that is suitable for harmonic studies. The models have been validated and compared using simulations executed on MATLAB /Simulink environment.

Performance Comparison of Machine Learning Models for Concrete Compressive Strength Prediction.

This study explores the prediction of concrete compressive strength using machine learning models, aiming to overcome the time-consuming and complex nature of conventional methods. Four models-an artificial neural network (ANN), a multiple linear regression, a support vector machine, and a regression tree-are employed and compared for performance, using evaluation metrics such as mean absolute deviation, root mean square error, coefficient of correlation, and mean absolute percentage error. After preprocessing 1030 samples, the dataset is split into two subsets: 70% for training and 30% for testing. The ANN model, further divided into training, validation (15%), and testing (15%), outperforms others in accuracy and efficiency. This outcome streamlines compressive strength determination in the construction industry, saving time and simplifying the process.

Machine learning ensembles, neural network, hybrid and sparse regression approaches for weather based rainfed cotton yield forecast.

Cotton is a major economic crop predominantly cultivated under rainfed situations. The accurate prediction of cotton yield invariably helps farmers, industries, and policy makers. The final cotton yield is mostly determined by the weather patterns that prevail during the crop growing phase. Crop yield prediction with greater accuracy is possible due to the development of innovative technologies which analyses the bigdata with its high-performance computing abilities. Machine learning technologies can make yield prediction reasonable and faster and with greater flexibility than process based complex crop simulation models. The present study demonstrates the usability of ML algorithms for yield forecasting and facilitates the comparison of different models. The cotton yield was simulated by employing the weekly weather indices as inputs and the model performance was assessed by nRMSE, MAPE and EF values. Results show that stacked generalised ensemble model and artificial neural networks predicted the cotton yield with lower nRMSE, MAPE and higher efficiency compared to other models. Variable importance studies in LASSO and ENET model found minimum temperature and relative humidity as the main determinates of cotton yield in all districts. The models were ranked based these performance metrics in the order of Stacked generalised ensemble > ANN > PCA ANN > SMLR ANN > LASSO> ENET > SVM > PCA SMLR > SMLR SVM > SMLR. This study shows that stacked generalised ensembling and ANN method can be used for reliable yield forecasting at district or county level and helps stakeholders in timely decision-making.

Multimodal brain age prediction using machine learning: combining structural MRI and 5-HT2AR PET-derived features.

To better assess the pathology of neurodegenerative disorders and the efficacy of neuroprotective interventions, it is necessary to develop biomarkers that can accurately capture age-related biological changes in the human brain. Brain serotonin 2A receptors (5-HT2AR) show a particularly profound age-related decline and are also reduced in neurodegenerative disorders, such as Alzheimer's disease. This study investigates whether the decline in 5-HT2AR binding, measured in vivo using positron emission tomography (PET), can be used as a biomarker for brain aging. Specifically, we aim to (1) predict brain age using 5-HT2AR binding outcomes, (2) compare 5-HT2AR-based predictions of brain age to predictions based on gray matter (GM) volume, as determined with structural magnetic resonance imaging (MRI), and (3) investigate whether combining 5-HT2AR and GM volume data improves prediction. We used PET and MR images from 209 healthy individuals aged between 18 and 85years (mean = 38, std = 18) and estimated 5-HT2AR binding and GM volume for 14 cortical and subcortical regions. Different machine learning algorithms were applied to predict chronological age based on 5-HT2AR binding, GM volume, and the combined measures. The mean absolute error (MAE) and a cross-validation approach were used for evaluation and model comparison. We find that both the cerebral 5-HT2AR binding (mean MAE = 6.63years, std = 0.74years) and GM volume (mean MAE = 6.95years, std = 0.83years) predict chronological age accurately. Combining the two measures improves the prediction further (mean MAE = 5.54years, std = 0.68). In conclusion, 5-HT2AR binding measured using PET might be useful for improving the quantification of a biomarker for brain aging.

Exploring the associations between structural and functional aspects of social relationships and cognition in very old age

Objectives Few studies have examined the associations between social relationships and cognition in very old adults (80+ years). Moreover, limited research has considered structural and functional aspects of social relationships together when exploring their associations with cognition. Therefore, the present study aimed to investigate the associations between structural and functional social aspects and cognition in very old age. Method The sample comprised 518 two-wave participants of the NRW80+ Study, who showed no indication of dementia and an average age of 85 years at baseline (range: 80–97 years). They were assessed twice over an approximate 2-year interval. Multilevel linear growth curve analysis was employed to model cognitive development over very old age and to examine the associations between structural and functional aspects of social relationships and cognitive development. It was controlled for demographic and health characteristics. Results A larger social network size (ß = 0.05, 95% CI [0.02, 0.08], p = 0.002) and frequent compared with occasional, seldom, and no contact with others (ß = 0.16, 95% CI [0.03, 0.28], p = 0.013) were associated with higher global cognition in the fully-controlled model. Model comparisons further revealed that structural aspects better predicted global cognition in very old age compared to functional aspects (χ2(2) = 16.17, p < 0.001). Conclusion The findings underline the positive association between structural aspects of social relationships and global cognition in very old age. They also highlight the need for interventions promoting a socially active lifestyle to prevent cognitive decline in this vulnerable age group.

Meta-Feature-Based Traffic Accident Risk Prediction: A Novel Approach to Forecasting Severity and Incidence

This study aims to improve the accuracy of predicting the severity of traffic accidents by developing an innovative traffic accident risk prediction model—StackTrafficRiskPrediction. The model combines multidimensional data analysis including environmental factors, human factors, roadway characteristics, and accident-related meta-features. In the model comparison, the StackTrafficRiskPrediction model achieves an accuracy of 0.9613, 0.9069, and 0.7508 in predicting fatal, serious, and minor accidents, respectively, which significantly outperforms the traditional logistic regression model. In the experimental part, we analyzed the severity of traffic accidents under different age groups of drivers, driving experience, road conditions, light and weather conditions. The results showed that drivers between 31 and 50 years of age with 2 to 5 years of driving experience were more likely to be involved in serious crashes. In addition, it was found that drivers tend to adopt a more cautious driving style in poor road and weather conditions, which increases the margin of safety. In terms of model evaluation, the StackTrafficRiskPrediction model performs best in terms of accuracy, recall, and ROC–AUC values, but performs poorly in predicting small-sample categories. Our study also revealed limitations of the current methodology, such as the sample imbalance problem and the limitations of environmental and human factors in the study. Future research can overcome these limitations by collecting more diverse data, exploring a wider range of influencing factors, and applying more advanced data analysis techniques.

A Comparison of Markov and Mechanistic Models for Soil-Transmitted Helminth Prevalence Projections in the Context of Survey Design.

Globally, there are over 1 billion people infected with soil-transmitted helminths (STHs), mostly living in marginalized settings with inadequate sanitation in sub-Saharan Africa and Southeast Asia. The World Health Organization recommends an integrated approach to STH morbidity control through improved access to sanitation and hygiene education and the delivery of preventive chemotherapy (PC) to school-age children delivered through schools. Progress of STH control programs is currently estimated using a baseline (pre-PC) school-based prevalence survey and then monitored using periodical school-based prevalence surveys, known as Impact Assessment Surveys (IAS). We investigated whether integrating geostatistical methods with a Markov model or a mechanistic transmission model for projecting prevalence forward in time from baseline can improve IAS design strategies. To do this, we applied these 2 methods to prevalence data collected in Kenya, before evaluating and comparing their performance in accurately informing optimal survey design for a range of IAS sampling designs. We found that, although both approaches performed well, the mechanistic method more accurately projected prevalence over time and provided more accurate information for guiding survey design. Both methods performed less well in areas with persistent STH hotspots where prevalence did not decrease despite multiple rounds of PC. Our findings show that these methods can be useful tools for more efficient and accurate targeting of PC. The general framework built in this paper can also be used for projecting prevalence and informing survey design for other neglected tropical diseases.

Non-locality of mean scalar transport in two-dimensional Rayleigh–Taylor instability using the macroscopic forcing method

The importance of non-locality of mean scalar transport in two-dimensional Rayleigh–Taylor Instability (RTI) is investigated. The macroscopic forcing method is utilized to measure spatio-temporal moments of the eddy diffusivity kernel representing passive scalar transport in the ensemble averaged fields. Presented in this work are several studies assessing the importance of the higher-order moments of the eddy diffusivity, which contain information about non-locality, in models for RTI. First, it is demonstrated through a comparison of leading-order models that a purely local eddy diffusivity is insufficient to capture the mean field evolution of the mass fraction in RTI. Therefore, higher-order moments of the eddy diffusivity operator are not negligible. Models are then constructed by utilizing the measured higher-order moments. It is demonstrated that an explicit operator based on the Kramers–Moyal expansion of the eddy diffusivity kernel is insufficient. An implicit operator construction that matches the measured moments is shown to offer improvements relative to the local model in a converging fashion.

3D electromagnetic analytical model for radial-flux eddy-current couplers

PurposeThis paper aims to develop a new 3D analytical model in cylindrical coordinates to study radial flux eddy current couplers (RFECC) while considering the magnetic edge and 3D curvature effects, and the field reaction due to the induced currents.Design/methodology/approachThe analytical model is developed by combining two formulations. A magnetic scalar potential formulation in the air and the magnets regions and a current density formulation in the conductive region. The magnetic field and eddy currents expressions are obtained by solving the 3D Maxwell equations in 3D cylindrical coordinates with the variable separation method. The torque expression is derived from the field solution using the Maxwell stress tensor. In addition to 3D magnetic edge effects, the proposed model takes into account the reaction field effect due to the induced currents in the conducting part. To show the accuracy of the developed 3D analytical model, its results are compared to those from the 3D finite element simulation.FindingsThe obtained results prove the accuracy of the new developed 3D analytical model. The comparison of the 3D analytical model with the 2D simulation proves the strong magnetic edge effects impact (in the axial direction) in these devices which must be considered in the modelling. The new analytical model allows the magnetic edge effects consideration without any correction factor and also presents a good compromise between precision and computation time.Practical implicationsThe proposed 3D analytical model presents a considerably reduced computation time compared to 3D finite element simulation which makes it efficient as an accurate design and optimization tool for radial flux eddy current devices.Originality/valueA new analytical model in 3D cylindrical coordinates has been developed to find the electromagnetic torque in radial flux eddy current couplers. This model considers the magnetic edge effects, the 3D curvature effects and the field reaction (without correction factors) while improving the computation time.

Formulation of stress intensity factor under pure bending condition in multilayer pavements using numerical study and model tree approach

AbstractThe tensile deformation of top‐down cracks significantly influences pavement delamination. Accurately formulating the stress intensity factor (SIF) is crucial for assessing the stress/strain near crack tips. In the present study, a robust data mining method called M5´ is employed to formulate the SIF induced by pure bending in a multilayer pavement. A comprehensive database is generated through two‐dimensional finite element analysis to train the M5’ model. Several key parameters pertaining to asphalt and base layers are incorporated into the formulation. A comparative analysis is conducted to evaluate the performance of the developed model against traditional regression methods. A comparison of the developed model with other existing methods confirms its superior performance. Additionally, a parametric study validates the robustness of the developed model in capturing fundamental concepts of fracture mechanics. Finally, a sensitivity analysis is performed to determine the most critical parameters in the development of a predictive model for SIF.

Radiation Biological Toximetry Using Circulating Cell-Free DNA (cfDNA) for Rapid Radiation/Nuclear Triage.

Optimal triage biodosimetry would include risk stratification within minutes, and it would provide useful triage despite heterogeneous dosimetry, cytokine therapy, mixed radiation quality, race, and age. For regulatory approval, the U.S. Food and Drug Administration (FDA) Biodosimetry Guidance requires suitability for purpose and a validated species-independent mechanism. Circulating cell-free DNA (cfDNA) concentration assays may provide such triage information. To test this hypothesis, cfDNA concentrations were measured in unprocessed monkey plasma using a branched DNA (bDNA) technique with a laboratory developed test. Therefore, cfDNA concentration measurements are increasingly used in radiation oncology clinics to predict side effect risk. The cfDNA levels, along with hematopoietic parameters, were measured over a 7-day period in Rhesus macaques receiving total body radiation doses ranging from 1 to 6.5 Gy. Low-dose irradiation (0-2 Gy) was easily distinguished from high-dose whole-body exposures (5.5 and 6.5 Gy). Fold changes in cfDNA in the monkey model were comparable to those measured in a bone marrow transplant patient receiving a supralethal radiation dose, suggesting that the lethal threshold of cfDNA concentrations may be similar across species. Average cfDNA levels were 50 ± 40 ng/mL [±1 standard deviation (SD)] pre-irradiation, 120 ± 13 ng/mL at 1 Gy; 242 ± 71 ng/mL at 2 Gy; 607 ± 54 at 5.5 Gy; and 1585 ± 351 at 6.5 Gy (±1 SD). There was an exponential increase in cfDNA concentration with radiation dose. Comparison of the monkey model with the mouse model and the Guskova model, developed using Chernobyl responder data, further demonstrated correlation across species, supporting a similar mechanism of action. The test is available commercially in a Clinical Laboratory Improvement Amendments (CLIA) ready form in the U.S. and the European Union. The remaining challenges include developing methods for further simplification of specimen processing and assay evaluation, as well as more accurate calibration of the triage category with cfDNA concentration cutoffs.

Evaluation of Deep Learning-Based Non-Intrusive Thermal Load Monitoring

Non-Intrusive Load Monitoring (NILM), which provides sufficient load for the energy consumption of an entire building, has become crucial in improving the operation of energy systems. Although NILM can decompose overall energy consumption into individual electrical sub-loads, it struggles to estimate thermal-driven sub-loads such as occupants. Previous studies proposed Non-Intrusive Thermal Load Monitoring (NITLM), which disaggregates the overall thermal load into sub-loads; however, these studies evaluated only a single building. The results change for other buildings due to individual building factors, such as floor area, location, and occupancy patterns; thus, it is necessary to analyze how these factors affect the accuracy of disaggregation for accurate monitoring. In this paper, we conduct a fundamental evaluation of NITLM in various realistic office buildings to accurately disaggregate the overall thermal load into sub-loads, focusing on occupant thermal load. Through experiments, we introduce NITLM with deep learning models and evaluate these models using thermal load datasets. These thermal load datasets are generated by a building energy simulation, and its inputs for the simulation were derived from realistic data like HVAC on/off data. Such fundamental evaluation has not been done before, but insights obtained from the comparison of learning models are necessary and useful for improving learning models. Our experimental results shed light on the deep learning-based NITLM models for building-level efficient energy management systems.

Calculation and prediction of divertor detachment via impurity seeding by using 1D model

Abstract Achieving the divertor detachment helps to alleviate excessive heat load and sputtering problems on the target plates, thereby prolonging the lifetime of divertor components for fusion devices. In order to provide a rapid but relatively reliable prediction of plasma parameters along the flux tube for future devices design, a one-dimensional (1D) modeling code for the impurity seeded detached divertor operational point has been developed based on the Python language, which is a fluid model based on the previous work.[1] The experimental observation of the onset of divertor detachment by neon (Ne) and argon (Ar) seeding in EAST is well reproduced by using the 1D modeling code. The comparison of the 1D modeling and 2D simulation by SOLPS-ITER code for CFETR detachment operation with Ne and Ar seeding also shows good consistency. The predictions of the radiative power loss and requirement of the corresponding impurity concentration for achieving divertor detachment via different impurity seeding for the high heating power scenarios on EAST and CFETR Phase II by the 1D model are also presented. Based on the predictions, the optimized parameter space for divertor detachment operation on EAST and CFETR has also been determined. Such a simple but reliable 1D model can provide a reasonable parameter inputs for a detailed and accurate analysis by 2D or 3D modeling tools through rapid parameter scanning.

Mixed bioconvection of nanofluid of oxytactic bacteria through a porous cavity with inlet and outlet under periodic magnetic field using artificial intelligence based on LightGBM algorithm

The comprehension of microorganisms’ responses to magnetic field and fluid motion offers potential for the advancement of targeted drug delivery systems or medical interventions reliant on biological fluids. In this paper, the mixed bioconvection flow of nanofluid of oxytactic bacteria has been investigated through a porous cavity with inlet and outlet ports under the impact of periodic magnetic field. All the walls of the cavity are fixed at the constant high temperature. The proposed problem has been modeled first and then simulated using the finite element method. The computed computational fluid dynamics results have been analyzed for the several important controlling parameters. It is observed that there is at least 65% increment on heat transfer between the lowest and highest Peclet numbers. Further, regression analysis was conducted using the LightGBM algorithm. Careful parameter tuning was performed to avoid overfitting, ensuring that the model did not memorize the training data. According to the R2 values used for regression analysis, in the 18 datasets used for the performance metric comparison of the artificial intelligence model, a total of 18 targets were tried to be predicted for different Richardson number (0.1, 1, and 10) and Lewis Number (0.1, 1, and 10) values for each motile microorganisms, temperature and oxygen concentration values specified in the datasets, and a minimum accuracy of 95% and a maximum accuracy of 98% were obtained. These findings demonstrate the capability of the LightGBM algorithm to accurately predict the target variable within a high range of accuracy for the given datasets.

MCell4 with BioNetGen: A Monte Carlo simulator of rule-based reaction-diffusion systems with Python interface.

Biochemical signaling pathways in living cells are often highly organized into spatially segregated volumes, membranes, scaffolds, subcellular compartments, and organelles comprising small numbers of interacting molecules. At this level of granularity stochastic behavior dominates, well-mixed continuum approximations based on concentrations break down and a particle-based approach is more accurate and more efficient. We describe and validate a new version of the open-source MCell simulation program (MCell4), which supports generalized 3D Monte Carlo modeling of diffusion and chemical reaction of discrete molecules and macromolecular complexes in solution, on surfaces representing membranes, and combinations thereof. The main improvements in MCell4 compared to the previous versions, MCell3 and MCell3-R, include a Python interface and native BioNetGen reaction language (BNGL) support. MCell4's Python interface opens up completely new possibilities for interfacing with external simulators to allow creation of sophisticated event-driven multiscale/multiphysics simulations. The native BNGL support, implemented through a new open-source library libBNG (also introduced in this paper), provides the capability to run a given BNGL model spatially resolved in MCell4 and, with appropriate simplifying assumptions, also in the BioNetGen simulation environment, greatly accelerating and simplifying model validation and comparison.

Structural equation modeling for the effects of family dysfunctions and communication on perceived mental health status among under/graduate students in the U.S.

Many of the college students expressed experiencing moderate or severe psychological distress. In their emergence as adults, college students encounter significant periods of drastic change. Family functions play a crucial role in determining individuals' mental health status. The first objective of this study was to investigate whether family communication mediates the relationships between family dysfunctions and perceived mental health status among under/graduate students. The second objective was to investigate whether the mediation effects of family communication on the relationships between family dysfunctions and perceived mental health status differ by gender. The current study applied Minuchin's Structural Family Theory with under/graduate students (N = 348) in Ohio. Structural Equation Modeling was conducted to examine the effects of family dysfunctions and communication on perceived mental health status among under/graduate students. Disengaged family function significantly influenced family communication, which, in turn, significantly impacted perceived mental health (mediation effect). Enmeshed family function did not show a significant relationship with family communication, whereas family communication had a significant impact on perceived mental health (no mediation effect). The model comparison test indicated that there were differences between males and females: the structural weights of the current path models were consistent with the full models, and there was a significant difference in the effect of parents' marital status on perceived mental health, with males being significantly impacted in both models. Practitioners should recognize that improving positive communication skills within the family unit can be a realistic approach to supporting the perceived mental health of under/graduate students. Colleges/universities should offer regular education programs to promote an understanding of the relationship between family communication strategies and the mental health of under/graduate students.

Comparison of deep learning models for milk production forecasting at national scale

Forecasting of agricultural variables at national scale is essential for developing policies aimed at guaranteeing food security and increase the stability of the entire agri-food chain. However, this is often a challenging task due to the fact that data are usually sparse and long records are rarely available. This work aims at evaluating the potentiality of ten deep learning models for predicting monthly milk production in France, Germany, and Italy, using as input, climatic and economic variables from open datasets. The results indicate that deep learning models are a better alternative to traditional statistical approaches, providing robust results without requiring complex model architectures. Furthermore, the prominent autoregressive nature of milk production highlights the inability of environmental variables to capture external processes that influence the milk production (e.g., milk quota). However, the high accuracy achieved by the models lays the ground for possible application of deep learning models to accurately forecast agricultural variables at national scale with potential implications for the development of dairy insurance products, and risk management practices.

Comparison of Early Contrast Enhancement Models in Ultrafast Dynamic Contrast-Enhanced Magnetic Resonance Imaging of Prostate Cancer.

Tofts models have failed to produce reliable quantitative markers for prostate cancer. We examined the differences between prostate zones and lesion PI-RADS categories and grade group (GG) using regions of interest drawn in tumor and normal-appearing tissue for a two-compartment uptake (2CU) model (including plasma volume (vp), plasma flow (Fp), permeability surface area product (PS), plasma mean transit time (MTTp), capillary transit time (Tc), extraction fraction (E), and transfer constant (Ktrans)) and exponential (amplitude (A), arrival time (t0), and enhancement rate (α)), sigmoidal (amplitude (A0), center time relative to arrival time (A1 - T0), and slope (A2)), and empirical mathematical models, and time to peak (TTP) parameters fitted to high temporal resolution (1.695 s) DCE-MRI data. In 25 patients with 35 PI-RADS category 3 or higher tumors, we found Fp and α differed between peripheral and transition zones. Parameters Fp, MTTp, Tc, E, α, A1 - T0, and A2 and TTP all showed associations with PI-RADS categories and with GG in the PZ when normal-appearing regions were included in the non-cancer GG. PS and Ktrans were not associated with any PI-RADS category or GG. This pilot study suggests early enhancement parameters derived from ultrafast DCE-MRI may become markers of prostate cancer.

AI model for predicting adult cochlear implant candidacy using routine behavioral audiometry

ObjectiveTo describe an AI model to facilitate adult cochlear implant candidacy prediction based on basic demographical data and standard behavioral audiometry. MethodsA machine-learning approach using retrospective demographic and audiometric data to predict candidacy CNC word scores and AzBio sentence in quiet scores was performed at a tertiary academic center. Data for the model were derived from adults completing cochlear implant candidacy testing between January 2011 and March 2023. Comparison of the prediction model to other published prediction tools and benchmarks was performed. ResultsThe final dataset included 770 adults, encompassing 1045 AzBio entries, and 1373 CNC entries. Isophoneme scores and word recognition scores exhibited strongest importance to both the CNC and AzBio prediction models, followed by standard pure tone average and low-frequency pure tone average. The mean absolute difference between the predicted and actual score was 15 percentage points for AzBio sentences in quiet and 13 percentage points for CNC word scores, approximating anticipated test-retest constraints inherent to the variables incorporated into the model. Our final combined model achieved an accuracy of 87 % (sensitivity: 90 %; precision: 80 %). ConclusionWe present an adaptive AI model that predicts adult cochlear implant candidacy based on routine behavioral audiometric and basic demographical data. Implementation efforts include a public-facing online prediction tool and accompanying smartphone program, an embedded notification flag in the electronic medical record to alert providers of potential candidates, and a program to retrospectively engage past patients who may be eligible for cochlear implantation based on audiogram results.