Prediction Accuracy Of Method Research Articles

Analysis and Prediction of Epidemic Prevention and Control by Police Stations Based on Time Series

<p>It has been over two years since the outburst of the COVID-19 pandemic. Currently, China has entered into a normalization stage and police stations are still in the endeavor of improving their epidemic prevention and control measures. However, grassroots police stations are still backward in epidemic prevention and control, and lack of response measures for each period of the epidemic. This paper uses time series models to predict the epidemic trend and analyze the measures undertaken by the police stations. In the process of data pretreatment, this paper focuses on the data processing of the epidemic control period. Then the epidemic trend is predicted based on five different time series models and two different time intervals. The results indicate that the tertiary exponential smoothing prediction model with day as the interval is the best and accurate prediction method. According to the prediction model, it can be determined the current stage of the epidemic development by time points, so as to give targeted reference for the police stations. The basic idea in using various time series models is to predict the accumulated number of confirmed cases based on the existing data not only to help, guide and refine the existing epidemic measures but also offer suggestions for epidemic prevention and control by police stations in response to each period of the epidemic. Based on the findings exhaustive recommendations are proposed for real-time and targeted epidemic prevention and control by police administration.</p> <p> </p>

Drug-Protein Interactions Prediction Models Using Feature Selection and Classification Techniques.

Drug-Protein Interaction (DPI) identification is crucial in drug discovery. The high dimensionality of drug and protein features poses challenges for accurate interaction prediction, necessitating the use of computational techniques. Docking-based methods rely on 3D structures, while ligand-based methods have limitations such as reliance on known ligands and neglecting protein structure. Therefore, the preferred approach is the chemogenomics-based approach using machine learning, which considers both drug and protein characteristics for DPI prediction. In machine learning, feature selection plays a vital role in improving model performance, reducing overfitting, enhancing interpretability, and making the learning process more efficient. It helps extract meaningful patterns from drug and protein data while eliminating irrelevant or redundant information, resulting in more effective machine-learning models. On the other hand, classification is of great importance as it enables pattern recognition, decision-making, predictive modeling, anomaly detection, data exploration, and automation. It empowers machines to make accurate predictions and facilitates efficient decision-making in DPI prediction. For this research work, protein data was sourced from the KEGG database, while drug data was obtained from the DrugBank data machine-learning base. To address the issue of imbalanced Drug Protein Pairs (DPP), different balancing techniques like Random Over Sampling (ROS), Synthetic Minority Over-sampling Technique (SMOTE), and Adaptive SMOTE were employed. Given the large number of features associated with drugs and proteins, feature selection becomes necessary. Various feature selection methods were evaluated: Correlation, Information Gain (IG), Chi-Square (CS), and Relief. Multiple classification methods, including Support Vector Machines (SVM), Random Forest (RF), Adaboost, and Logistic Regression (LR), were used to predict DPI. Finally, this research identifies the best balancing, feature selection, and classification methods for accurate DPI prediction. This comprehensive approach aims to overcome the limitations of existing methods and provide more reliable and efficient predictions in drug-protein interaction studies.

Identification of pan-cancer/testis genes and validation of therapeutic targeting in triple-negative breast cancer: Lin28a-based and Siglece-based vaccination induces antitumor immunity and inhibits metastasis

BackgroundCancer–testis (CT) genes are targets for tumor antigen-specific immunotherapy given that their expression is normally restricted to the immune-privileged testis in healthy individuals with aberrant expression in tumor tissues. While...

Indoor Plant Soil-Plant Analysis Development (SPAD) Prediction Based on Multispectral Indices and Soil Electroconductivity: A Deep Learning Approach

Leaf Soil-Plant Analysis Development (SPAD) prediction is a crucial measure of plant health and is essential for optimizing indoor plant management. The deep learning methods offer advanced tools for precise evaluations but their adaptation to the heterogeneous indoor plant ecosystem presents distinct challenges. This study assesses how accurately deep neural network (DNN) predicts SPAD values in leaves on indoor plants when compared to well-established machine learning techniques, including Random Forest (RF) and Extreme Gradient Boosting (XGB). The covariates for prediction were based on low-cost multispectral and soil electro-conductivity (EC) sensors, enabling a non-destructive sensing approach. The study also strongly emphasized multicollinearity analysis quantified by the Variance Inflation Factor (VIF) and two independent indices, as well as its effect on prediction accuracy using deep and machine learning methods. DNN resulted in higher accuracy to RF and XGB, also performing better using filtered data after multicollinearity analysis based on the coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE) (R2 = 0.589, RMSE = 11.68, MAE = 9.52) in comparison to using all input covariates (R2 = 0.476, RMSE = 12.90, MAE = 10.94). Overall, DNN was proven as a more accurate prediction method than the conventional machine learning approach for the prediction of leaf SPAD values in indoor plants, despite using heterogenous plant types and input covariates.

Development and validation of a weight-loss predictor to assist weight loss management

This study aims to develop and validate a modeling framework to predict long-term weight change on the basis of self-reported weight data. The aim is to enable focusing resources of health systems on individuals that are at risk of not achieving their goals in weight loss interventions, which would help both health professionals and the individuals in weight loss management. The weight loss prediction models were built on 327 participants, aged 21–78, from a Finnish weight coaching cohort, with at least 9 months of self-reported follow-up weight data during weight loss intervention. With these data, we used six machine learning methods to predict weight loss after 9 months and selected the best performing models for implementation as modeling framework. We trained the models to predict either three classes of weight change (weight loss, insufficient weight loss, weight gain) or five classes (high/moderate/insufficient weight loss, high/low weight gain). Finally, the prediction accuracy was validated with an independent cohort of overweight UK adults (n = 184). Of the six tested modeling approaches, logistic regression performed the best. Most three-class prediction models achieved prediction accuracy of > 50% already with half a month of data and up to 97% with 8 months. The five-class prediction models achieved accuracies from 39% (0.5 months) to 89% (8 months). Our approach provides an accurate prediction method for long-term weight loss, with potential for easier and more efficient management of weight loss interventions in the future. A web application is available: https://elolab.shinyapps.io/WeightChangePredictor/.The trial is registered at clinicaltrials.gov/ct2/show/NCT04019249 (Clinical Trials Identifier NCT04019249), first posted on 15/07/2019.

Analysis of prediction methods for non-equilibrium internal flow condensation heat transfer

Due to the climate crisis, there is a need to improve the performance of energy conversion systems and increase the use of renewable energy sources. In this context, accurate performance prediction methods for heat exchangers, including condensers, are of crucial importance. In the heat transfer process in a condenser including desuperheating and subcooling, both sensible and latent heat transfer may occur simultaneously due to thermal non-equilibrium effects. Liquid condensate may form close to the vapor saturation point of the bulk flow in the desuperheating region, while superheated vapor may penetrate beyond the liquid saturation point far into the two-phase region, and gas bubbles and subcooled mixture may exist in the subcooled region. These are complex phenomena that to a large extent affect the performance of the condenser. This paper presents an analysis of the prediction methods for non-equilibrium condensation heat transfer. First, the heat transfer mechanisms of non-equilibrium condensation are discussed. Next, state-of-the-art prediction methods developed for the target heat transfer processes are analyzed. Moreover, a database containing all data available in the open literature is built to evaluate the predictive performance of the prediction methods. Finally, a new prediction method is proposed for the subcooled condensation based on the identified heat transfer mechanisms. The results suggest that the Jacob correlation has the best predictive performance, followed by the Kondou correlation. The mean absolute percentage errors of both the Jacob and Kondou prediction methods are lower than 15 %, and more than 85 % of the experimental data points are predicted with a deviation within ± 30 %. The new prediction method proposed for the subcooled condensation is in good agreement with the experimental data, with a mean absolute percentage error of 12 %.

Machine learning-based assessment of hygrothermal aging performance in CFRP-aluminum alloy adhesive bonded structures

ABSTRACT Adhesive bonding is an effective technique for connecting fiber reinforced composites (FRPs) and plays a crucial role in determining load transfer capability and overall structural safety. However, long-term exposure to temperature and moisture can degrade adhesive joints, highlighting the significant importance of accurately predicting their durability when designing reliable adhesive structures. This study aims to address the industry’s demand and academic interest by proposing a convenient and accurate prediction method to evaluate the durability of adhesive structures subject to complex stresses. Laboratory accelerated aging experiments were conducted on butt joints (experiencing pure normal stress) and shear joints (experiencing pure shear stress). Quasi-static tensile tests and adhesive FTIR tests were performed to establish a dataset of adhesive functional group absorbance and failure strength under single stress conditions. To compare and assess the predictive performance of different models for aging behaviors, Artificial Neural Network (ANN), Support Vector Regression (SVR) and Random Forest (RF) models were selected. The model hyperparameters were further optimized using the Genetic Algorithm (GA). The results demonstrated that the ANN model exhibited superior correlation and accuracy compared to the SVR, and RF. The characteristic importance analysis revealed that the hydrolysis and post-curing of adhesives were the primary factors affecting joint performance under hygrothermal environments. By incorporating environmental degradation factors and defining initial failure criteria, the CZM constitutive model was modified and a predictive method was established for assessing the hygrothermal aging behavior of CFRP/aluminum alloy bonding structures. This method effectively eliminates the effect the intricate aging process (e.g. aging time, aging conditions) and establishes a direct correlation between the molecular structure of adhesive and the macroscopic mechanical properties of the bonded joint, which serves as a valuable reference and guidance for the safety design of automotive composite material bonding structures throughout their entire lifespan.

CellTICS: an explainable neural network for cell-type identification and interpretation based on single-cell RNA-seq data.

Identifying cell types is crucial for understanding the functional units of an organism. Machine learning has shown promising performance in identifying cell types, but many existing methods lack biological significance due to poor interpretability. However, it is of the utmost importance to understand what makes cells share the same function and form a specific cell type, motivating us to propose a biologically interpretable method. CellTICS prioritizes marker genes with cell-type-specific expression, using a hierarchy of biological pathways for neural network construction, and applying a multi-predictive-layer strategy to predict cell and sub-cell types. CellTICS usually outperforms existing methods in prediction accuracy. Moreover, CellTICS can reveal pathways that define a cell type or a cell type under specific physiological conditions, such as disease or aging. The nonlinear nature of neural networks enables us to identify many novel pathways. Interestingly, some of the pathways identified by CellTICS exhibit differential expression "variability" rather than differential expression across cell types, indicating that expression stochasticity within a pathway could be an important feature characteristic of a cell type. Overall, CellTICS provides a biologically interpretable method for identifying and characterizing cell types, shedding light on the underlying pathways that define cellular heterogeneity and its role in organismal function. CellTICS is available at https://github.com/qyyin0516/CellTICS.

SLAPP: Subgraph-level attention-based performance prediction for deep learning models

The intricacy of the Deep Learning (DL) landscape, brimming with a variety of models, applications, and platforms, poses considerable challenges for the optimal design, optimization, or selection of suitable DL models. One promising avenue to address this challenge is the development of accurate performance prediction methods. However, existing methods reveal critical limitations. Operator-level methods, proficient at predicting the performance of individual operators, often neglect broader graph features, which results in inaccuracies in full network performance predictions. On the contrary, graph-level methods excel in overall network prediction by leveraging these graph features but lack the ability to predict the performance of individual operators. To bridge these gaps, we propose SLAPP, a novel subgraph-level performance prediction method. Central to SLAPP is an innovative variant of Graph Neural Networks (GNNs) that we developed, named the Edge Aware Graph Attention Network (EAGAT). This specially designed GNN enables superior encoding of both node and edge features. Through this approach, SLAPP effectively captures both graph and operator features, thereby providing precise performance predictions for individual operators and entire networks. Moreover, we introduce a mixed loss design with dynamic weight adjustment to reconcile the predictive accuracy between individual operators and entire networks. In our experimental evaluation, SLAPP consistently outperforms traditional approaches in prediction accuracy, including the ability to handle unseen models effectively. Moreover, when compared to existing research, our method demonstrates a superior predictive performance across multiple DL models.

Enhancing groundwater level prediction accuracy at a daily scale through combined machine learning and physics-based modeling

Study regionYongding River alluvial proluvial fan, a part of North China Plain with intense groundwater withdrawals. Study focusThe objective of this study is to enhance the accuracy of groundwater level (GWL) prediction at a daily scale by combining short-term memory (LSTM) and physics-based (PB) models. Two types of LSTM models were developed: LSTM-H-ORI, which predicts GWL, and LSTM-dH-ORI, which predicts simulated errors of the PB model. These models were subsequently refined into LSTM-H-IMP and LSTM-dH-IMP by incorporating the outputs from the PB model. To investigate the influence of groundwater flow processes, LSTM-H-IMP was further enhanced by considering the time lag properties of precipitation, referred to as LSTM-H-IMP-LAG. New hydrological insights for the regionSignificant improvements were found for LSTM-H-IMP when the performance of LSTM-H-ORI was poor (Nash–Sutcliffe Efficiency Coefficient, NSE<0) or the performance of PB model was not bad (NSE≥0). And the prediction accuracy was improved for over 67% of wells in this case. When the performance of PB model was medium and poor (NSE≤0.6), the improvement of LSTM-dH-IMP was more effective, leading to a prediction accuracy enhancement for over 77% of wells. Additionally, LSTM-H-IMP-LAG exhibited further improvement, with an average NSE increase of 0.1. This study provides scientific methods for accurate prediction of GWL at a daily time scale and the combined application of LSTM and PB models.

Multi-view spatiotemporal learning for traffic forecasting

The spatiotemporal learning has been one of the most burning research topics in traffic data analytics for traffic forecasting, which is instrumental in developing the intelligent transportation systems. Capturing accurately and efficiently complex spatiotemporal dependencies of the road network is a critical prerequisite of the traffic spatiotemporal forecasting. This study proposes a multi-view spatiotemporal learning (MVSTL) framework, which combines the fast parallel learning (FPL) and the serial learning (SL) to juggle the receptive field with the fitting capability of the model. The FPL is proposed to capture the spatiotemporal correlations synchronously and efficiently by reducing the number of parameters significantly. As a supplement to the FPL, the SL, which extracts the spatial and temporal features serially, can expand the spatial/temporal receptive fields and avoid information overwriting. The evaluation of the MVSTL is carried out by employing four sets of traffic data from the LA freeway system, and the experiment results reveal that the MVSTL is able to perform better than the advanced methods in prediction accuracy.

Measurement of Non-Contact Uterine Fundus in Second and Third Trimester of Pregnancy Using Machine Learning

Regular pregnancy checks provide good data points for pregnant women. The information obtained is expected to have an impact on the health of the baby from the womb until the baby is born. Fetal monitoring and examination have developed significantly in terms of the use of technology as a tool for control and examination. In harmony with digitalization being promoted by the government through digital-based health services. On the other hand, apart from causing things to change the order of life, including behavior and habits, one of the things is the role of technology, which is increasingly vital in its initial nature as a tool for human life, including in the health sector. One of the things that can be integrated with technology is pregnancy monitoring for pregnant women, especially measuring the symphysis-fundal height. Professionals should have an accurate prediction method for making diagnoses during prenatal care, so they can detect them as early as possible. This can be assisted by detecting the size of the symphysis-fundal height using digital images. This study took data from 98 respondents with a distribution of 4 data collection points. Of the 98 data, there were 10 invalid data because manual measurements failed to take place. This study took data from 98 respondents with a distribution of 4 data collection points. This study is the best accuracy at the Banjaran midwife location, with a success rate of 93%. Further research is needed with a larger sample of pregnant women so that machine learning has an adequate trial to become a data-based maternal health. The implication of these findings is that machine learning has not been able to predict gestational age with precision, so it must be given more samples so that machine learning can continue to be trained in predicting and calculating after the end of the second and third trimesters.

A rapid performance prediction method for Two-Phase liquid metal MHD generators based on Quasi-One-Dimensional model

Liquid metal magnetohydrodynamic (LMMHD) power generation is a novel and highly efficient technology with broad application prospects. In our previous study, a novel LMMHD enhanced Closed-Brayton-Cycle power generation system for hypersonic vehicles was proposed, and its excellent power generation performance was demonstrated by thermodynamic analysis. The most important component of LMMHD system is the power generation channel, and the rapid and accurate performance prediction and optimization design methods of which are the prerequisites for further research. In this study, a quasi-one-dimensional modeling method for LMMHD power generation channels based on a two-fluid model is developed. The performance variation laws with the geometric structure, magnetic field and boundary conditions, as well as the potential loss mechanisms are investigated. Results indicate that the most optimal or appropriate design can be achieved by optimizing the channel geometry and magnetic field, such as the optimal aspect ratio of 1 and the optimal Hartmann number of 2000. But it is difficult to balance the power output and efficiency through the boundary condition optimization. The gas–liquid slip can lead to approximately 20–40% of energy losses, making the elimination and suppression of slip effects a significant research focus in the relevant field.

Matrix decomposition methods for accurate water saturation prediction in Canadian oil-sands by LF-NMR T2 measurements

This article presents a novel method for quantifying water saturation in oil-sand reservoirs by employing 1D low-field nuclear magnetic resonance (LF-NMR) spin-spin relaxation and bulk density measurements as indicators of pore volume variations. One of the challenges in accurately determining the volumes of bitumen and water in oil-sands is the effective separation of their overlapping T2 signals, attributed to similar spin-spin relaxation decay times and diffusive coupling. Conventional methods require deconvolution of T2 peaks and or experimentation to determine T2 cutoff values, differentiating between bitumen and water signals, notably capillary and clay-bound water. In contrast, our approach predicts the proportion of water by utilizing matrix decomposition methods to compress the T2 relaxation distribution and extract significant components. These components subsequently train the regression model, facilitating the accurate estimation of relative water saturation percentages.The NMR dataset was obtained by benchtop LF-NMR T2 measurements from 82 oil-sand samples, with preserved bitumen and water saturations at both reservoir and ambient temperatures (6 °C and 25 °C), yielding 164 observations. We examined four matrix decomposition methods, including principal component analysis, its variation integrating a kernel function, canonical correlation analysis, and partial least squares regression. X-ray CT measurements and Dean-Stark extraction ascertained the respective sample bulk densities and fluid-solid volume proportions.The PCA model prediction statistics (RMSE = 0.86%, R2 = 0.84), indicate its application can be extended for saturation prediction from NMR and bulk density well logs. Moreover, we underscore the importance of incorporating bulk density measurements and establish the statistical and physical correlations between these measurements and NMR T2 relaxation, providing insights into the approach's efficacy and causality.

Field observations and long short-term memory modeling of spectral wave evolution at living shorelines in Chesapeake Bay, USA

Living shorelines as a nature-based solution for climate change adaptation were constructed in many places around the world. The success of this type of projects requires long-term monitoring for adaptive management. The paper presents a novel framework leveraging scientific machine learning methods for accurate and rapid prediction of long-term hydrodynamic forcing impacting living shorelines using short-term measurements of water levels and wind waves in the largest estuary in the U.S. Different from existing data-driven wave prediction models focusing on significant wave heights, this study is focused on the prediction of wave energy spectra in shallow water using winds and tides as the input feature and short-term measurements of wave spectra and water depths as the label. Long Short-Term Memory (LSTM) models were developed using four-month wave measurements in the stormy seasons to predict integral wave parameters and energy spectra for multiple years. The developed models accurately predicted wave heights, peak periods, and energy spectra around the living shorelines, capturing complex wave dynamics, such as wave generation by wind, nonlinear wave-wave interactions, and depth-limited wave breaking in the shallow water of a large estuary. The validated models were then used to determine the long-term wave forcing impacting the living shorelines based on the modeled wave characteristics and spectra. Model results show that the surrogate models utilizing LSTM to predict wave spectra in the frequency domain enable long-term predictions of spectral wave evolution with a minimal computational cost. Our findings provide valuable insights into the efficacy of living shorelines in attenuating wave energy and demonstrate the utility of this approach in assessing the effectiveness of such living shoreline structures.

Learning a World Model With Multitimescale Memory Augmentation.

Model-based reinforcement learning (RL) is regarded as a promising approach to tackle the challenges that hinder model-free RL. The success of model-based RL hinges critically on the quality of the predicted dynamic models. However, for many real-world tasks involving high-dimensional state spaces, current dynamics prediction models show poor performance in long-term prediction. To that end, we propose a novel two-branch neural network architecture with multi-timescale memory augmentation to handle long-term and short-term memory differently. Specifically, we follow previous works to introduce a recurrent neural network architecture to encode history observation sequences into latent space, characterizing the long-term memory of agents. Different from previous works, we view the most recent observations as the short-term memory of agents and employ them to directly reconstruct the next frame to avoid compounding error. This is achieved by introducing a self-supervised optical flow prediction structure to model the action-conditional feature transformation at pixel level. The reconstructed observation is finally augmented by the long-term memory to ensure semantic consistency. Experimental results show that our approach is able to generate visually-realistic long-term predictions in DeepMind maze navigation games, and outperforms the prevalent state-of-the-art methods in prediction accuracy by a large margin. Furthermore, we also evaluate the usefulness of our world model by using the predicted frames to drive an imagination-augmented exploration strategy to improve the model-free RL controller.

A Two-Stage Power Distribution Scheme of Multiple Wind Farms Participating in Primary Frequency Regulation

With the increasing penetration rate of renewable energy power generation (RPG) in power system, improving the frequency regulation ability of RPG has become a general trend. The premise of effective participation of RPG in primary frequency regulation (PFR) is to reasonably design the power distribution scheme for RPG devices. In order to effectively evaluate the frequency support capability of wind power generation system, the system frequency response (SFR) model under the joint action of synchronous generators, wind farms (WF) and low frequency load shedding device is established in this paper. According to the time-domain solution of this mathematical model, the influence of the output power of the WF on SFR is mathematically explained. At the same time, the influence of time delay on SFR is further demonstrated. In order to maximize the release of the rotor kinetic energy of the wind turbine and ensure the safe operation of the wind turbine when designing the power distribution scheme, an accurate prediction method of wind turbine rotor speed based on piecewise fitting and Bernoulli equation is proposed. Based on the established SFR model and rotor speed prediction method, a two-stage power distribution scheme for multiple WFs during PFR is proposed, which comprehensively considers the SFR requirements and different power output capabilities of multiple WFs. The effectiveness and superiority of the proposed power distribution scheme are validated with revised IEEE 39-bus test system and revised PG&E 69-bus test system.

Hysteresis characteristics prediction method of amorphous materials based on static Jiles-Atherton hysteresis model and Maxwell's equation

Establishing an accurate hysteresis prediction method is crucial for reliable numerical analysis, especially for the growing demand in industrial applications. Existing hysteresis prediction methods, unfortunately, suffer from high cost, time consumption, and inaccuracy, particularly for those under high frequency. For that, this paper proposes a dynamic Jiles-Atherton hysteresis method with practicality and effectiveness. Firstly, an analytical model is derived to predict the eddy current magnetic field strength under high-frequency excitation using Maxwell's equation. Secondly, a modified static hysteresis characteristics prediction approach considering the skin effect is proposed based on the static Jiles-Atherton hysteresis model. Finally, a dynamic Jiles-Atherton hysteresis method (DJAM) considering the influence of skin effect on hysteresis characteristics is presented under the statistical theory of losses. The comparison between the simulated and measured results not only confirms the accuracy and efficiency of the DJAM but also indicates that its prediction accuracy is superior to the traditional method, particularly for those under high frequency.

A multi-variable multi-step Seq2seq networks for the state of charge estimation of lithium-ion battery

Due to the complexity and changeable of lithium-ion batteries, we propose a multi-variable and multi-step Temporal neural network to cover this task. Specially, a novel multi-step training strategy is applied to deal with long time sequences, and multi-variables is added to supervise the prediction. In addition, the Seq2seq net with long short term memory (LSTM) modules is employed to improve the prediction accuracy in various lithium-ion battery operating environments of −10 °C, 10 °C, 25 °C and 40 °C. The A123 lithium-ion batteries’ database collected by University of Maryland for the real scenarios has been used in the training process. The ablation studies indicate that the multi-step module is necessary under the root mean square error, mean absolute error and mean absolute percentage error index. The prediction error of comparison experiment under our proposed algorithms is 0.45 %@ root mean square error (RMSE), 0.30 %@ mean absolute error (MAE) and 1.11 %@ mean absolute percentage error (MAPE) at 10 °C. Our study introduces a novel and accurate SOC prediction method, which can provide accurate SOC estimates in a wide range of scenarios, emphasizing its novelty and relevance in addressing real-world challenges.

Prediction of fatigue lifetime by the strain energy density for filled polychloroprene during thermo-oxidative ageing

During long-term environmental exposure, polychloroprene undergoes ageing. Throughout the process of thermo-oxidation, chemical degradation leads to crosslinking in the polychloroprene macromolecular network, inducing a significant change in mechanical properties. Among the various consequences of ageing on the properties, the reduction of fatigue lifetime is still poorly understood and needs more accurate prediction methods. This study presents a methodology to relate change in fatigue lifetime loss and network crosslinking during ageing. The material investigated is a carbon black-filled polychloroprene, subjected to more than 25 ageing conditions. Changes in network crosslink density, mechanical strain energy, and fatigue lifetime is experimentally measured. In the present study, a novel prediction strategy is proposed: it relies on the network strain energy density and convincingly describes fatigue loss. As the approach potentially extends to three-dimensional and/or multiaxial problems, it holds promise for addressing ageing-related issues in the analysis of elastomeric parts.