Model Predictions Research Articles

Study of a mechano-electrochemical model: a numerical and experimental approach

The mechanochemical effects of elasticity and plasticity are introduced into a peridynamic (PD) corrosion model. A PD equation that couples the mechanics and kinetics of electrochemistry is proposed for the first time. This clarifies the influence of the mechanical load in governing the occurrence of diffusion during corrosion. The phase changes experienced by a material are used to characterize the movement of the corrosion boundary, and the concentration of material points is used to characterize the degree and extent of damage caused to the chosen metal. Immersion tests and in situ electrochemical tests are carried out, and the results are used to explore the effects of stress and/or load on the kinetics of environment-induced damage to an aluminum alloy. The model predictions are in good agreement with the experimental observations. The results of this study demonstrate that a coupled mechano-electrochemical PD corrosion model can capture corrosion-induced damage and can be used to study damage propagation upon exposure to an aggressive environment.

The strengthened role of new predictors of Indian Ocean Dipole (IOD) during the recent decades of weakened ENSO-IOD relationship

Indian Ocean Dipole (IOD) is the major interannual ocean-atmosphere interaction phenomena in the Indian Ocean (IO) and is influenced by external forcing such as El Nino Southern Oscillation (ENSO). Meanwhile, its co-occurrence and stronger relationship with ENSO have decreased during recent decades. The IOD variability also reduced after 2000, accompanied by a shift of the western pole to central IO extending further southward. The study reports that the boreal fall (SON, September, October, November) season IOD has an intensified relationship with the previous winter Subtropical Indian Ocean dipole (SIOD) and spring season equatorial north tropical Atlantic (NTA) SST anomalies., while the ENSO relationship is reduced from its pre-2000 value. It is found that the persistent warming (cooling) in the western side of positive (negative) SIOD during the previous winter induces easterly (westerly) wind anomalies in the equatorial IO during the following summer and fall, leading to positive (negative) IOD events. These IOD events have the western pole shifted to south-central IO instead of the canonical northwestern warming. The spring season NTA SST anomalies induce stronger summer season circulation and SST gradient in the equatorial Pacific, like ENSO. During the SON season, this pattern is associated with cooling and easterly wind anomalies in the tropical eastern IO and IOD. These two patterns explain the major mode of IOD variability after 2000. While the IOD associated with NTA have co-occurring ENSO during summer and fall, the SIOD-induced IOD events are independent of ENSO in the Pacific. Thus, these two predictors provide long-lead predictability (2-3 seasons ahead) of IOD for both ENSO co-occurring and non-ENSO IOD events. However, the IOD predictability of seasonal prediction models mainly depends on the ENSO-IOD relationship, resulting in reduced IOD skills for many of them after 2000. The models such as COLA-CCSM4, which has improved skill have stronger than observed ENSO-IOD relationship than the pre-2000 period. A linear regression model including SIOD and NTA SST indices of the previous winter and spring season respectively as predictors simulates IOD with a skill of around 0.65 during the recent period, indicating the necessity of seasonal prediction models to capture the variability in the southern IO and NTA and their teleconnections for better prediction of IOD in the recent period of reduced ENSO skill.

Improving the Statistical Representation of Tropical Cyclone In-Storm Sea Surface Temperature Cooling

Abstract This study uses fixed buoy time series to create an algorithm for sea surface temperature (SST) cooling underneath a tropical cyclone (TC) inner core. To build predictive equations, SST cooling is first related to single variable predictors such as the SST before storm arrival, ocean heat content (OHC), mixed layer depth, sea surface salinity and stratification, storm intensity, storm translation speed, and latitude. Of all the single variable predictors, initial SST before storm arrival explains the greatest amount of variance for the change in SST during storm passage. Using a combination of predictors, we created nonlinear predictive equations for SST cooling. In general, the best predictive equations have four predictors and are built with knowledge about the prestorm ocean structure (e.g., OHC), storm intensity (e.g., minimum sea level pressure), initial SST values before storm arrival, and latitude. The best-performing SST cooling equations are broken up into two ocean regimes: when the ocean heat content is less than 60 kJ cm−2 (greater spread in SST cooling values) and when the ocean heat content is greater than 60 kJ cm−2 (SST cooling is always less than 2°C), which demonstrates the importance of the prestorm oceanic thermal structure on the in-storm SST value. The new equations are compared to what is currently used in a statistical–dynamical model. Overall, since the ocean providing the latent heat and sensible heat fluxes necessary for TC intensification, the results highlight the importance for consistently obtaining accurate in-storm upper-oceanic thermal structure for accurate TC intensity forecasts. Significance Statement The ocean provides the heat and moisture necessary for tropical cyclone (TC) intensification. Since the heat and moisture transfer depend on the sea surface temperature (SST), we create statistical equations for the prediction of SST underneath the storm. The variables we use combine the initial SST before the storm arrives, the upper-ocean thermal structure, and the strength and translation speed of the storm. The predictive equations for SST are evaluated for how well they improve TC intensity forecasts. The best-performing equations can be used for prediction in operational statistical models, which can aid intensity forecasts.

Posterior Representations for Bayesian Context Trees: Sampling, Estimation and Convergence

We revisit the Bayesian Context Trees (BCT) modelling framework for discrete time series, which was recently found to be very effective in numerous tasks including model selection, estimation and prediction. A novel representation of the induced posterior distribution on model space is derived in terms of a simple branching process, and several consequences of this are explored in theory and in practice. First, it is shown that the branching process representation leads to a simple variable-dimensional Monte Carlo sampler for the joint posterior distribution on models and parameters, which can efficiently produce independent samples. This sampler is found to be more efficient than earlier MCMC samplers for the same tasks. Then, the branching process representation is used to establish the asymptotic consistency of the BCT posterior, including the derivation of an almost-sure convergence rate. Finally, an extensive study is carried out on the performance of the induced Bayesian entropy estimator. Its utility is illustrated through both simulation experiments and real-world applications, where it is found to outperform several state-of-the-art methods.

Decision tree-based approach to extrapolate life cycle inventory data of manufacturing processes

Life cycle assessment (LCA) plays a crucial role in green manufacturing to uncover the critical aspects for alleviating the environmental burdens due to manufacturing processes. However, the scarcity of life cycle inventory (LCI) data for the manufacturing processes is a considerable challenge. This paper proposes a novel approach to extrapolate LCI data of manufacturing processes. Taking advantage of LCI data in the Ecoinvent datasets, decision tree-based supervised machine learning models, namely decision tree, random forest, gradient boosting, and adaptive boosting, have been developed to extrapolate the data of GHG emissions, i.e., carbon dioxide, nitrous oxide, methane, and water vapor. Initially, a correlation analysis was conducted to derive the most influential factors on GHG quantities resulting from manufacturing activities. First, the collected data have been preprocessed and split into train and test sets (70% and 30%, respectively). Second, a five-fold cross-validation method was applied to tune the hyperparameters of the models. Then, the models were re-trained using the best hyperparameters and evaluated using the test set. The results reveal that the Gradient Boosting model has a superior predictive performance for extrapolating the GHG emission data, with average coefficients of determination (R2) on the test set <0.95. Moreover, the model predictions involve relatively low values of the average root mean squared error and an average mean percentage of error on the test set. The correlation and feature importance analyses emphasized that the workpiece material and manufacturing technology have a considerable effect on natural resource consumption, i.e., energy, material, and water inflows into the process. Meanwhile, energy consumption, water usage, and raw aluminum depletion were the most influential factors in GHG emissions. Eventually, a case study to extrapolate the inflows and the outflows for new manufacturing activities has been conducted using the validated models. The proposed GraBoost model provides a computational supplementary approach to estimate and extrapolate the GHG emissions for different manufacturing processes when LCI data are incomplete or don't exist within LCI databases.

A Robust Approach to Malware Detection through Ensemble learning

Abstract: With the proliferation of digital threats in today's interconnected world, the need for efficient and effective automated malware detection systems has become paramount. Traditional signature-based methods often fail to keep pace with the evolving landscape of malware, necessitating the development of more sophisticated techniques. In this paper, we propose a robust approach to automated malware detection leveraging ensemble learning techniques. Ensemble learning, which combines the predictions of multiple base models, has demonstrated remarkable success in various domains, including cybersecurity. Our approach harnesses the diversity of ensemble methods to enhance the detection accuracy and robustness against adversarial attacks. By integrating diverse base classifiers such as decision trees, random forests, support vector machines, and neural networks, our ensemble model learns to effectively discriminate between benign and malicious software samples. Furthermore, we introduce novel feature engineering strategies tailored to capture the intricate characteristics of malware. These features encompass a wide range of attributes, including static file properties, dynamic behavioural patterns, and frequency-based representations. Leveraging these rich feature sets, our ensemble model can generalize well across different types of malware families and variants.

Deepfake Audio Detection System

Abstract: To ensure the authenticity of audio material, it is important to establish reliable detection tools that can trace the spread and use of deepfake technologies. This study focuses on the methodology of deepfake audio identification via Mel Frequency Cepstral Coefficients (MFCC) as features with Random Forest as a classifier. By extracting MFCC features from audio clips and using them in a Random Forest model, one can learn unique spectral properties that can help distinguish deepfakes from authentic audio. The Random Forest algorithm, famous for its quality of being able to work well in an ensemble learning paradigm, is utilized to identify patterns that are representative of deepfake manipulation. To ensure the efficiency and reliability of this method, it was tested on a large number of different data sets that included both genuine and fake voice samples. To ensure robustness and generalization, cross-validation techniques are employed, restricting model predictions to the range of 0 to 1 and providing informative error messages for effective diagnosis. Thus, it is an important scientific study helping to develop and strengthen methods for identifying and eliminating threats in the area of artificial sound activity.

Use of explainable machine learning models in blast load prediction

The effects of blast waves and their consequent damage to structures have been an increasingly popular research topic in the past decade. Various methods are used in blast load prediction on structures, including experimental, semi-empirical and numerical approaches. However, there is a demand for developing time-efficient predictive methods that help various professionals, such as engineering practitioners and first responders, in their regular activities. Machine learning (ML), as a subset of artificial intelligence (AI), has gained significant attention in various engineering applications over the past decades for its ability to aid decision-making. However, a critical gap exists in the comprehensive analysis of these ML models' transparency and interpretability levels. This paper presents a novel approach to employing explainable machine learning (XML) to predict the blast loads generated by high explosives. The primary objective is to assess the validity of utilising a machine learning (ML) framework for blast load predictions, examining the scientific consistency of the model's prediction process. This research utilises XML to interpret the ML model-based blast load prediction framework, marking the first documented use of XML in published literature for blast load prediction. This research study utilised three ML models, namely a) Decision Tree, b) Random Forest, and c) Extreme Gradient Boost (XGB) models, to predict blast loads on rigid structures. Validated numerical model results were used to generate the dataset to train the ML models, and the respective dataset consists of approximately 600 independent blast events. The models were evaluated for their predictability using error metrics: adjusted coefficient of determination (adjusted R2), root mean square error (RMSE), and mean absolute error (MAE). The overall ML model performance evaluation showed that the best-performing model, XGB, could make predictions with 98 % accuracy compared to validated numerical predictions. The XML explanations for the XGB models showcased that the blast load prediction procedure of the models aligns with the blast physics principles, ensuring the credibility of the developed ML models. In summary, these XML predictive models can be considered accurate and reliable for blast load prediction on structures.

Mathematical modeling of rare earth element separation in electrodialysis with adjacent anion exchange membranes and ethylenediaminetetraacetic acid as chelating agent

This research delves into the effective use of electrodialysis for the separation of rare earth elements (REEs), specifically separating dysprosium (Dy) from praseodymium (Pr) and neodymium (Nd). A robust mathematical model based on the extended Nernst-Planck equation is introduced, simulating the process within a configuration that includes two adjacent anion exchange membranes. The model integrates aspects such as feed equilibrium, ion exchange within the membrane, and overall ion flux. Validation of the model's predictability was conducted through Chi-squared tests and root mean square error (RMSE) calculations, affirming its capability to accurately predict ion concentrations across different compartments. The study examines essential parameters such as applied voltage, rinse solution concentration, and feed concentration, assessing their impacts on separation performance and energy efficiency. Results indicate that higher voltages above 8 V, while speeding up separation, detrimentally impact energy use. It also highlights a critical balance in rinse solution concentration; lower concentrations below 0.05 mol/L enhance energy efficiency but may undercut separation efficacy due to early depletion. A linear correlation between the necessary rinse concentration and feed concentration was established, with higher feed concentrations demonstrating reduced specific energy consumption, thus enhancing overall efficiency. However, challenges remain in current efficiency due to the independent migration of SO42− ions in this specific setup. The findings advocate exploring alternative configurations, like alternating cation and anion exchange membranes, to optimize both environmental and economic aspects of REE separation. This study provides valuable insights and recommendations for refining electrodialysis systems in REE processing, contributing to sustainable and cost-effective electrodialysis systems.

Surprise-minimization as a solution to the structural credit assignment problem.

The structural credit assignment problem arises when the causal structure between actions and subsequent outcomes is hidden from direct observation. To solve this problem and enable goal-directed behavior, an agent has to infer structure and form a representation thereof. In the scope of this study, we investigate a possible solution in the human brain. We recorded behavioral and electrophysiological data from human participants in a novel variant of the bandit task, where multiple actions lead to multiple outcomes. Crucially, the mapping between actions and outcomes was hidden and not instructed to the participants. Human choice behavior revealed clear hallmarks of credit assignment and learning. Moreover, a computational model which formalizes action selection as the competition between multiple representations of the hidden structure was fit to account for participants data. Starting in a state of uncertainty about the correct representation, the central mechanism of this model is the arbitration of action control towards the representation which minimizes surprise about outcomes. Crucially, single-trial latent-variable analysis reveals that the neural patterns clearly support central quantitative predictions of this surprise minimization model. The results suggest that neural activity is not only related to reinforcement learning under correct as well as incorrect task representations but also reflects central mechanisms of credit assignment and behavioral arbitration.

Stochastic Optimization of the Management Schedule of Korean Pine Plantations

Korean pine is one of the most important tree species in northeastern China, where Korean pine plantations produce timber and edible seeds. Often, seeds create more income than timber. Predicting the timber and cone yields of alternative management schedules of the plantations involves uncertainty because the future climatic conditions for tree growth and cone production are unknown. This study developed a simulation model that generates stochastic variation around the predictions of tree growth and cone yield models, allowing the forest manager to seek cutting schedules that maximize the expected amounts of timber or cones, or the expected economic profit, under uncertain future states of nature. Stochastic analysis also facilitates management optimizations for different risk attitudes. The differential evolution algorithm and the developed stochastic simulation model were used to optimize the management of planted Korean pine. Timber and cone yields of a management schedule were calculated under 100 different scenarios for tree growth and cone production. When the growth and cone yield scenarios were stationary (no temporal trends), the optimal management schedules were similar to those of deterministic optimization. The benefits of stochastic optimization increased when it was assumed that the tree growth scenarios may contain climate-change-induced trends. Non-stationary growth variation led to shorter optimal rotation lengths than stationary growth variation. Increasing risk tolerance shortened optimal rotations.

Power deposition in the He bath induced by the Z gradient coil in the 11.7 T Iseult magnet: theory versus measurements

Gradient-magnet interactions increase with higher field magnets and stronger gradient coils. Perhaps the most sensitive aspect is magnet quench, which can be induced by a loss of superconductivity of the main coil caused by a rise of the temperature of the He bath with gradient activity. Predicting power depositions thereby can be a very valuable tool to avoid dangerous frequency zones but also eventually correct design flaws. In this work, we report model predictions compared to measurements of power deposition in the He bath of the Iseult 11.7 T magnet for the Z gradient coil axis.

Redox potentials in ionic liquids: Anomalous behavior?

Redox potentials depend on the nature of the solvent/electrolyte through the solvation energies of the ionic solute species. For concentrated electrolytes, ion solvation may deviate significantly from the Born model predictions due to ion pairing and correlation effects. Recently, Ghorai and Matyushov [J. Phys. Chem. B 124, 3754-3769 (2020)] predicted, on the basis of linear response theory, an anomalous trend in the solvation energies of room temperature ionic liquids, with deviations of hundreds of kJ/mol from the Born model for certain size solutes/ions. In this work, we computationally evaluate ionic solvation energies in the prototypical ionic liquid, 1-butyl-3-methylimidazolium tetrafluoroborate (BMIM/BF4), to further explore this behavior and benchmark several of the approximations utilized in the solvation energy predictions. For comparison, we additionally compute solvation energies within acetonitrile and molten NaCl salt to illustrate the limiting behavior of purely dipolar and ionic solvents. We find that the overscreening effect, which results from the inherent charge oscillations of the ionic liquid, is substantially reduced in magnitude due to screening from the dipoles of the molecular ions. Therefore, for the molten NaCl salt, for which the ions do not have permanent dipoles, modulation of ionic solvation energies from the overscreening effect is most significant. The conclusion is that ionic liquids do indeed exhibit unique solvation behavior due to peak(s) in the electrical susceptibility caused by the ion shell structure; redox potential shifts for BMIM/BF4 are of more modest order ∼0.1V, but may be larger for other ionic liquids that approach molten salt behavior.

Development of an Elementary Reaction-Based Kinetic Model to Predict the Aqueous-Phase Fate of Organic Compounds Induced by Reactive Free Radicals.

ConspectusAqueous-phase free radicals such as reactive oxygen, halogen, and nitrogen species play important roles in the fate of organic compounds in the aqueous-phase advanced water treatment processes and natural aquatic environments under sunlight irradiation. Predicting the fate of organic compounds in aqueous-phase advanced water treatment processes and natural aquatic environments necessitates understanding the kinetics and reaction mechanisms of initial reactions of free radicals with structurally diverse organic compounds and other reactions. Researchers developed conventional predictive models based on experimentally measured transformation products and determined reaction rate constants by fitting with the time-dependent concentration profiles of species due to difficulties in their measurements of unstable intermediates. However, the empirical treatment of lumped reaction mechanisms had a model prediction limitation with respect to the specific parent compound's fate. We use ab initio and density functional theory quantum chemical computations, numerical solutions of ordinary differential equations, and validation of the outcomes of the model with experiments. Sensitivity analysis of reaction rate constants and concentration profiles enables us to identify an important elementary reaction in formating the transformation product. Such predictive elementary reaction-based kinetics models can be used to screen organic compounds in water and predict their potentially toxic transformation products for a specific experimental investigation.Over the past decade, we determined linear free energy relationships (LFERs) that bridge the kinetic and thermochemical properties of reactive oxygen species such as hydroxyl radicals (HO•), peroxyl radicals (ROO•), and singlet oxygen (1O2); reactive halogen species such as chlorine radicals (Cl•) and bromine radicals (Br•); reactive nitrogen species (NO2•); and carbonate radicals (CO3•-). We used literature-reported experimental rate constants as kinetic information. We considered the theoretically calculated aqueous-phase free energy of activation or reaction to be a kinetic or a thermochemical property, and obtained via validated ab initio or density functional theory-based quantum chemical computations using explicit and implicit solvation models. We determined rate-determining reaction mechanisms involved in reactions by observing robust LFERs. The general accuracy of LFERs to predict aqueous-phase rate constants was within a difference of a factor of 2-5 from experimental values.We developed elementary reaction-based kinetic models and predicted the fate of acetone induced by HO• in an advanced water treatment process and methionine by photochemically produced reactive intermediates in sunlit fresh waters. We provided mechanistic insight into peroxyl radical reaction mechanisms and critical roles in the degradation of acetone and the formation of transformation products. We highlighted different roles of triplet excited states of two surrogate CDOMs, 1O2, and HO•, in methionine degradation. Predicted transformation products were compared to those obtained via benchtop experiments to validate our elementary reaction-based kinetic models. Predicting the reactivities of reactive halogen and nitrogen species implicates our understanding of the formation of potentially toxic halogen- and nitrogen-containing transformation products during water treatment processes and in natural aquatic environments.

The impact of dose rate on responses of human lens epithelial cells to ionizing irradiation

The knowledge on responses of human lens epithelial cells (HLECs) to ionizing radiation exposure is important to understand mechanisms of radiation cataracts that are of concern in the field of radiation protection and radiation therapy. However, biological effects in HLECs following protracted exposure have not yet fully been explored. Here, we investigated the temporal kinetics of γ-H2AX foci as a marker for DNA double-strand breaks (DSBs) and cell survival in HLECs after exposure to photon beams at various dose rates (i.e., 150 kVp X-rays at 1.82, 0.1, and 0.033 Gy/min, and 137Cs γ-rays at 0.00461 Gy/min (27.7 cGy/h) and 0.00081 Gy/min (4.9 cGy/h)), compared to those in human lung fibroblasts (WI-38). In parallel, we quantified the recovery for DSBs and cell survival using a biophysical model. The study revealed that HLECs have a lower DSB repair rate than WI-38 cells. There is no significant impact of dose rate on cell survival in both cell lines in the dose-rate range of 0.033–1.82 Gy/min. In contrast, the experimental residual γ-H2AX foci showed inverse dose rate effects (IDREs) compared to the model prediction, highlighting the importance of the IDREs in evaluating radiation effects on the ocular lens.

Thermodynamic performance of SiC-enhanced MicroPCM backfill based on response surface methodology

The functional backfilling in mining achieves synergistic co-mining of deep-earth deposits-geothermal, and the development of functional filling materials with suitable phase change parameters and excellent physicochemical properties is imminent. This study aims to investigate the optimal composition of phase change energy storage backfill material based on MicroPCM. To address the issues of reduced compressive strength and specific heat capacity of MicroPCM-backfill, silicon carbide (SiC) is introduced to prepare MPSC-backfill. Box-Behnken Design (BBD) is proposed to design and analyze according to four factors (ash-sand ratio, slurry concentration, MicroPCM, SiC), three levels (compressive strength, thermal conductivity, specific heat capacity) and three center points. Based on the Response Surface Methodology, the multivariate quadratic polynomial model was used to perform multiple linear regression and binomial fitting analysis to explore the optimal ratio. Combined with SEM-EDS (Scanning Electron Microscope with Energy-Dispersive X-ray Spectroscopy) analysis, the following conclusions were drawn. SiC exhibits an overall enhancement of the compressive strength of the backfill, and the thermal conductivity tends to increase almost linearly. The specific heat capacity grows in a small amount of silicon carbide doping, but excessive amounts lead to a decrease, with an overall limited impact. The addition of SiC does not affect the phase change ability of MicroPCM. And within a certain range of added quantities, it optimizes specific heat capacity, thermal conductivity, and compressive strength. Experimental validation based on the obtained optimal composition shows errors within a 5% margin, indicating the response model's good accuracy and predictability. High and low-temperature cyclic tests confirm the long-term thermal energy storage stability of MPSC-backfill during phase change cycles.

Unlocking the power of AI models: exploring protein folding prediction through comparative analysis.

Protein structure determination has made progress with the aid of deep learning models, enabling the prediction of protein folding from protein sequences. However, obtaining accurate predictions becomes essential in certain cases where the protein structure remains undescribed. This is particularly challenging when dealing with rare, diverse structures and complex sample preparation. Different metrics assess prediction reliability and offer insights into result strength, providing a comprehensive understanding of protein structure by combining different models. In a previous study, two proteins named ARM58 and ARM56 were investigated. These proteins contain four domains of unknown function and are present in Leishmania spp. ARM refers to an antimony resistance marker. The study's main objective is to assess the accuracy of the model's predictions, thereby providing insights into the complexities and supporting metrics underlying these findings. The analysis also extends to the comparison of predictions obtained from other species and organisms. Notably, one of these proteins shares an ortholog with Trypanosoma cruzi and Trypanosoma brucei, leading further significance to our analysis. This attempt underscored the importance of evaluating the diverse outputs from deep learning models, facilitating comparisons across different organisms and proteins. This becomes particularly pertinent in cases where no previous structural information is available.

Comparing Two Deep-Learning Models in Discrete Choice Analysis of Customers’ Mobile Plan Preferences

A discrete choice experiment (DCE) was used to investigate students’ preferences for mobile phone plans at a South African university. Upon obtaining the data, this study compares the predictive performance of two machine-learning models for discrete choice analysis and makes recommendations for model selection. Using concepts from blocked fractional factorial designs, a locally optimal DCE was created for the choice sets. This contrasts with alternative ways that, in practice, could be more difficult, especially when there is a large number of attributes. The call rate, data speed, customer service, premiums, and network coverage were the features considered. A total of 180 respondents were chosen from the student population using a two-stage sample approach, and data were gathered through face-to-face interviews. In this study, two deep-learning models are examined to analyze the data, which are the artificial neural network (ANN) and the extreme gradient boosting (XGBoost) models. Root mean square error (RMSE) and mean absolute error (MAE) are used to assess the model fitness, while accuracy, precision, recall and F1 score were used to compare the models’ performance. The results showed that XGBoost performs better compared to ANN in model fitness and prediction. Thus, the use of the XGBoost deep-learning model in choice preference modeling is therefore encouraged.

Making Predictions Using Poorly Identified Mathematical Models

Many commonly used mathematical models in the field of mathematical biology involve challenges of parameter non-identifiability. Practical non-identifiability, where the quality and quantity of data does not provide sufficiently precise parameter estimates is often encountered, even with relatively simple models. In particular, the situation where some parameters are identifiable and others are not is often encountered. In this work we apply a recent likelihood-based workflow, called Profile-Wise Analysis (PWA), to non-identifiable models for the first time. The PWA workflow addresses identifiability, parameter estimation, and prediction in a unified framework that is simple to implement and interpret. Previous implementations of the workflow have dealt with idealised identifiable problems only. In this study we illustrate how the PWA workflow can be applied to both structurally non-identifiable and practically non-identifiable models in the context of simple population growth models. Dealing with simple mathematical models allows us to present the PWA workflow in a didactic, self-contained document that can be studied together with relatively straightforward Julia code provided on GitHub. Working with simple mathematical models allows the PWA workflow prediction intervals to be compared with gold standard full likelihood prediction intervals. Together, our examples illustrate how the PWA workflow provides us with a systematic way of dealing with non-identifiability, especially compared to other approaches, such as seeking ad hoc parameter combinations, or simply setting parameter values to some arbitrary default value. Importantly, we show that the PWA workflow provides insight into the commonly-encountered situation where some parameters are identifiable and others are not, allowing us to explore how uncertainty in some parameters, and combinations of parameters, regardless of their identifiability status, influences model predictions in a way that is insightful and interpretable.

A general hyperelastic model for rubber-like materials incorporating strain-rate and temperature

A comprehensive hyperelastic model that precisely forecasts the mechanical characteristics of materials with rubber-like qualities is presented. This model relies on the well-established five-parameter Mooney-Rivlin model, which is then extended to incorporate strain rate dependent and temperature dependent term. To validate its accuracy, experimental data from ethylene propylene diene monomer (EPDM) rubber materials was utilized to compare with the model prediction. Hyperelastic stress-strain curves were collected from a variety of materials that were subjected to varying temperatures and strain rates to improve the model’s applicability. Its prediction results are compared against the collected experimental data, resulting in consistent and reliable outcomes. The simplified form of the model not only establishes an effective framework for characterizing and predicting the mechanical response of materials that resemble rubber under different working conditions but makes the coding and implementation of finite element analysis easier.