Adequate Prediction Research Articles

Examining the sensitivity of simulated EnMAP data for estimating chlorophyll-a and total suspended solids in inland waters

Our study investigates the capability of the environmental mapping and analysis program (EnMAP) scenes simulated using the EnMAP end-to-end simulator software (EeteS) based on the AISA Eagle airborne data to predict chlorophyll-a (Chl-a) and total suspended solids (TSS) as two of the most crucial water quality indicators. Three machine learning (ML) approaches (principal component regression(PCR), partial least square regression (PLSR) and random forest (RF)) were employed to establish links between the simulated image spectra and the above-mentioned water attributes of the samples collected from several inland water reservoirs within the southern part of the Czech Republic. Airborne hyperspectral images were also used to develop a model to compare its performance with models developed based on the simulated EnMAP data. Adequate prediction accuracy was obtained for both Chl-a (R2 = 0.89, RMSE = 43.06 μg/L, and Lin’s concordance correlation coefficient (LCCC) = 0.91) and TSS (R2 = 0.91, RMSE = 17.53 mg/L, and LCCC = 0.94), which were close enough to those obtained from the airborne hyperspectral images. Chl-a and TSS correlated with the wavelengths around 550 nm and 700 to 750 nm of the red and near-infrared (NIR) regions. In addition, the spatial distribution maps derived from the simulated EnMAP were comparable to those obtained from the AISA Eagle airborne data. Overall, it can be concluded that the simulated EnMAP image successfully and reliably predicted and spatially mapped the selected biophysical properties of the small inland water bodies.

SCORE2 cardiovascular risk prediction models in an ethnic and socioeconomic diverse population in the Netherlands: an external validation study.

Socioeconomic status and ethnicity are not explicitly incorporated as risk factors in the four SCORE2 cardiovascular disease (CVD) risk models developed for country-wide implementation across Europe (low, moderate, high and very-high model). The aim of this study was to evaluate the performance of the four SCORE2 CVD risk prediction models in an ethnic and socioeconomic diverse population in the Netherlands. The SCORE2 CVD risk models were externally validated in socioeconomic and ethnic (by country of origin) subgroups, from a population-based cohort in the Netherlands, with GP, hospital and registry data. In total 155,000 individuals, between 40 and 70 years old in the study period from 2007 to 2020 and without previous CVD or diabetes were included. Variables (age, sex, smoking status, blood pressure, cholesterol) and outcome first CVD event (stroke, myocardial infarction, CVD death) were consistent with SCORE2. 6966 CVD events were observed, versus 5495 events predicted by the CVD low-risk model (intended for use in the Netherlands). Relative underprediction was similar in men and women (observed/predicted (OE-ratio), 1.3 and 1.2 in men and women, respectively). Underprediction was larger in low socioeconomic subgroups of the overall study population (OE-ratio 1.5 and 1.6 in men and women, respectively), and comparable in Dutch and the combined "other ethnicities" low socioeconomic subgroups. Underprediction in the Surinamese subgroup was largest (OE-ratio 1.9, in men and women), particularly in the low socioeconomic Surinamese subgroups (OE-ratio 2.5 and 2.1 in men and women). In the subgroups with underprediction in the low-risk model, the intermediate or high-risk SCORE2 models showed improved OE-ratios. Discrimination showed moderate performance in all subgroups and the four SCORE2 models, with C-statistics between 0.65 and 0.72, similar to the SCORE2 model development study. The SCORE 2 CVD risk model for low-risk countries (as the Netherlands are) was found to underpredict CVD risk, particularly in low socioeconomic and Surinamese ethnic subgroups. Including socioeconomic status and ethnicity as predictors in CVD risk models and implementing CVD risk adjustment within countries is desirable for adequate CVD risk prediction and counselling. Leiden University Medical Centre and Leiden University.

Adoption of Deep Neural Network Model for the Prediction of Rain Attenuation

Weather, rain, and atmospheric forces cause attenuation in communication satellites and devices, which affects the quality of calls and internet access. Communication signals can seriously be affected by rainfall and various research has entailed developing adequate rainfall prediction models that can accurately predict rain attenuation. Predicting rainfall attenuation from measured data is a challenging process and several rainfall attenuation models exist, but their estimates are not close to actual measured rainfall attenuation data. This paper predicts rain attenuation using a Deep Neural Network (DNN) and modelling process to propose a probabilistic model that gives better results. The proposed model was built using information from the International Telecommunications Union Radiocommunication for propagation statistics and prediction techniques needed for the construction of Earth-space telecommunication systems. Rainfall data spanning a period of seven years (January 2014 to December 2020) in Oyo State were collected from Nigerian Meteorological Agency (NIMET). Results showed that the proposed model was able to predict well even with new data (2021) and this was used to show how well the DNN model can perform using rainfall features and elimination complex mathematical models to describe the mappings between rainfall attenuation and rainfall data. This is also evident from the performance graph which shows an RMSE of 0.21 and a Loss of 0.021 at the 100 epochs. The results also reveal that the proposed model results observed against predicted, performed relatively well. From the performance metrics, it shows a value of 0.7172 for Pearson’s linear correlation, showing a strong similarity of predicted versus observed. The mode efficiency index nt on the other hand, shows a value of 0.1744 and gives near acceptable results. In conclusion, the research showed that the Deep Neural Network model was able to predict well compared to the ITU-R model.

Ammonia void fraction in smooth tubes at different gravity orientation

Ammonia is an environment-friendly natural refrigerant and working fluid with negligible global warming and ozone depletion. Nowadays, ammonia is widely used in industrial refrigeration systems, heat pumps and thermal control systems, including heat pipes, loop heat pipes and two-phase mechanically pumped loops. In perspective, “green” ammonia is considered an energy carrier and carbon-free fuel. In two-phase systems, the void fraction is one of the critical factors determining the frictional and local pressure losses, boiling and condensation heat transfer and phase separation. The experimental study of ammonia void fraction in tubes with inner diameters of 8.5 mm and 15 mm is proposed in this paper. The quick-closing valve method has been chosen. Three flow configurations are considered: vertical downward, vertical upward and horizontal. The tests were provided for the two-phase ammonia at a mass velocity of 50…500 kg·m−2s−1, saturation pressure of 15…30 bar, and vapour quality of 0.05…0.30. One hundred and five points have been tested. No existing single correlation predicts the ammonia void fraction well at the considered boundary conditions for at least two orientations in the gravity field. A new single correlation has been proposed based on the Rouhani drift model with the tube inclination angle-dependent distribution parameter allowing the adequate prediction of void fraction in a tube with ammonia flow at horizontal, upward and downward orientations.

An Investigation into the Multi-Pass Radial-Mode Micro Abrasive Air Jet Turning of Fused-Silica Rods

With the increased requirement of miniaturization structures on hard and brittle substrates, micro abrasive air jet turning technology has become a promising machining technology for manufacturing miniaturization structures. This paper presents an investigation on multi-pass radial-mode micro abrasive air jet turning of fused silica. The effect of machining parameters on the depth of cut and materials removal rate was analyzed. The waviness and striation formation mechanisms were studied. It was found that increasing the number of passes significantly increases the depth of cut and materials removal rate. When the surface speed is reduced, the depth of cut and materials removal rate significantly increases. The waviness may be attributed to the intersection and overlap of the adjoint jet footprint. The crossed striations have been found, which may lead to the elimination of the striation. A predictive model of the material removal rate was developed using the dimensional analysis method. The model was assessed and shown to be able to give adequate predictions.

An analysis of the sensitivity of cutting force coefficients and their influence on the variability of stability diagrams

Process forces are an integral characteristic for the evaluation of machining operations, which can be calculated using, for instance, empirical models. An adequate prediction is essential, especially as it provides input data for subsequent models, e.g., for stability analysis of milling processes. However, the calibration of force model coefficients is not unambiguous and may have multiple local optima, which can significantly affect the accuracy of the approximation of the cutting forces. In this context, the selection of experiments used for calibration is crucial to obtain adequate results. In this paper, a systematic analysis of the selection of the required calibration experiments and the influence of the resulting parametrization of the coefficients on the predicted milling forces and the stability limit was conducted.Based on the results, designs of calibration experiments could be identified, with which the influence of varying undeformed chip thicknesses could be represented adequately. By applying these force model parameterizations for stability analysis, an improvement in the prediction of stability limits was achieved.

Vestibular Migraine versus Méniere's Disease: Diagnostic Utility of Electrocochleography.

Objectives: The diagnostic criteria for vestibular migraine (VM) and Méniere’s disease (MD) present an important overlap, which leads to a difficult diagnosis in patients presenting with headache, vertigo, hearing loss, ear fullness, and tinnitus. The objective of our study is to determine whether the area-under-the-curve ratio of the summating potentials (SP) and action potentials (AP) curves on electrocochleography (ECoG) helps differentiate VM from MD with or without the use of the well-established clinical criteria. Method: A retrospective review of patients filling either VM or MD criteria was undertaken between September 2015 and December 2018. All patients underwent ECoG before the introduction of anti-migraine therapy. The prediction of symptom improvement between the clinical criteria and ECoG results was compared by using the Vertigo Symptom Scale. Results: In total, 119 patients were included. An overlap of 36% exists between patients filling VM and MD criteria. Clinical criteria alone did not demonstrate a significant prediction of symptom response to anti-migraine therapy (VM 83%, MD 51%; p = 0.10). However, ECoG results alone did demonstrate adequate prediction (VM 94%, MD 32%; p < 0.001). A negative ECoG result combined with the clinical criteria of VM (100% symptom improvement) was shown to be more predictive of treatment response when compared to clinical criteria alone (83% symptom improvement) (p = 0.017). Finally, when used in patients filling both the VM and MD criteria (VMMD), ECoG was able to predict symptom improvement, thus better differentiating both diseases (normal ECoG: 95%, abnormal ECoG 29%; p < 0.001). Conclusion: Combining VM criteria with normal ECoG using the AUC ratio seems superior in predicting adequate symptom improvement than VM criteria alone.

UAS-based imaging for prediction of chickpea crop biophysical parameters and yield

Chickpea (Cicer arietinum L.) is a key legume crop grown in many semi-arid areas. Traditionally, chickpea is a rainfed spring crop, but in certain countries it has become an irrigated crop. The main objective of this study was to evaluate the ability of Unmanned Aerial Systems (UAS) imaging platform with an integrated RGB camera to provide estimations of leaf area index (LAI), biomass, and yield for chickpea during the irrigation period. Two field trials were conducted in 2019 and 2020, in which chickpea plants were subjected to five and six irrigation regimes, respectively. Eight vegetation indexes (VIs) and three morphological parameters were estimated from the RGB images. In parallel, biomass was determined, LAI was measured manually, and yield was determined at full maturity. In total, 294 plant samples were acquired and analyzed over the two years. Firstly, each of the VIs and morphological parameters were correlated separately against the two biophysical parameters and yield. Then, all the VIs and morphological parameters were analyzed together, and two statistical models, partial least squares regression (PLS-R) and support vector machine (SVM); were used to predict biomass and LAI. The yield was predicted using multi-linear regression (MLR). When each index or morphological parameter was analyzed separately, plant height and some of the VIs provided adequate predictions of the biophysical parameters in 2019 (R2 values ≥ 0.50) but failed (R2 values ≤ 0.25) in 2020. The integration of the VIs with the morphological parameters and the use of PLS-R and SVM models increased the accuracy level for both biophysical parameters (R2 ranged from 0.31 to 0.96) and mitigated the lack of consistency between the years. The SVM model was superior to the PLS-R model in both biophysical parameters. The R2 values for the combined 2019 and 2020 biomass model increased, at the model-testing stage, from 0.62 to 0.96 and the RMSE values dropped from 1778 to 490 kg ha−1. The ability of the SVM model to estimate chickpea biomass and LAI can provide convenient support for different management decisions, including timing and amount of irrigation and harvest date.

Improving Dementia Screening in India with Telemedicine and Artificial Intelligence: A Scalable Solution to Address Clinical Gaps in Care

AbstractBackgroundOver 4.4 million people live with Alzheimer’s and related dementia (ADRD) in India. There is significant clinical service gap of 90% and only 1 in 10 patients affected with ADRD in India receives screening, diagnosis, treatment, and care. Combining telemedicine with an “Expert system,” can help evaluate, screen, and triage patients with ADRD and reduce the time to diagnosis. We demonstrate early results from a feasibility study that uses state‐of‐the‐art Artificial Intelligence (AI), data analytics, and a HIPAA‐compliant Telehealth platform to screen patients over 65 years in India for ADRD.MethodWe integrated AI‐based platform and a standardized cognitive test into the telemedicine‐based primary‐care workflow. A primary care provider (PCP) screened every consenting patient aged 65 years and older on a telemedicine platform in India. The screening used an algorithm‐based intake tool, baseline Mini‐Mental Status Exam (MMSE), and clinical judgment of the PCP. We compared the final screening assessment generated by AI with the MMSE evaluation and clinical judgment of PCP for nine patients.ResultsNine patients (4 females, 5 males) in age group of 65‐90 years consented and enrolled in this initial study. Using the intake data as inputs, the AI model predicted MCI in 5 out of 9 patients (55%) and mild ADRD in 4 out of 9 patients (44%). The clinical severity from the tele‐encounter was rated as mild ADRD in 3 out of 9 patients (33%) and no ADRD/Mild Cognitive Impairment in 6 out of 9 patients (66%). Neuropsychological testing with the MMSE demonstrated scores from 24‐30 in 7 out of 9 patients (77%) and &lt;24 in 2 out of 9 patients (22%). The AI model accurately predicted the severity in 6 out of 9 patients (66.6%) of cases.ConclusionIt is feasible to combine AI and telemedicine to improve ADRD screening and clinical service gaps for healthcare delivery in India. AI severity models also demonstrated adequate predictability to triage patients who have cognitive impairment using digital intake data. The accuracy of the AI prediction models may be further improved with supplemental electronic medical data and translating the intake to a local language.

The PBPK LeiCNS-PK3.0 framework predicts Nirmatrelvir (but not Remdesivir or Molnupiravir) to achieve effective concentrations against SARS-CoV-2 in human brain cells

SARS-CoV-2 was shown to infect and persist in the human brain cells for up to 230 days, highlighting the need to treat the brain viral load. The CNS disposition of the antiCOVID-19 drugs: Remdesivir, Molnupiravir, and Nirmatrelvir, remains, however, unexplored. Here, we assessed the human brain pharmacokinetic profile (PK) against the EC90 values of the antiCOVID-19 drugs to predict drugs with favorable brain PK against the delta and the omicron variants. We also evaluated the intracellular PK of GS443902 and EIDD2061, the active metabolites of Remdesivir and Molnupiravir, respectively. Towards this, we applied LeiCNS-PK3.0, the physiologically based pharmacokinetic framework with demonstrated adequate predictions of human CNS PK. Under the recommended dosing regimens, the predicted brain extracellular fluid PK of only Nirmatrelvir was above the variants’ EC90. The intracellular levels of GS443902 and EIDD2061 were below the intracellular EC90. Summarizing, our model recommends Nirmatrelvir as the promising candidate for (pre)clinical studies investigating the CNS efficacy of antiCOVID-19 drugs.

중학생의 수학 교과역량에 영향을 미치는 언어 관련 개인, 가정, 수업 변인의 영향력 탐색

Objectives In this study, the influence of various predictive factors was analyzed through artificial neural network model in order to explore the influence of language-related personal, family, and class variables on middle school students' mathematics competency. Structural relationships were analyzed by constructing a model with the selected major variables as predictors and mediator, and the mathematics competency as dependent variables. Methods For this purpose, the self-report questionnaire data of 2,235 middle school students in the 3rd year(2019) of the D-educational longitudinal study and the scores of the Korean language and mathematics competency tests were used for the analysis. An artificial neural network model was constructed that predicts mathematics competency based on language-related individual, family, and class variables that can predict mathematics subject competency and Korean competency. Structural equation model analysis was conducted by setting the structural relationship between language-related variables and mathematics competency as a research model. Results The purposes of this study were to specify variables related to convergence competency, and to analyze the structural relationships among them. The purposes of this study were to specify variables related to convergence competency, and to analyze the structural relationships among them. Conclusions As a result of checking the correlation coefficient between the training data and the predicted value in the process of building the artificial neural network model, the average of the error was 0.7, and the correlation coefficient was . 68, indicating adequate prediction performance. As a result of confirming the structural relationship between language-related variables and Korean competency and mathematics competency, individual, family, and class variables directly or indirectly had a significant positive effect on mathematics subject competency.

An Intelligent Adequate-Fitting Prediction Method of Coastal Tunnel Rock Deformation Based on the Effective Rank of Hidden Layer

The intelligent prediction of surrounding rock deformation is of great significance for guiding the design and construction of tunnel projects in coastal areas. The deformation of tunnels in coastal areas is more complex than that of the ground, and the risk of encountering adverse geological conditions is greater. The traditional tunnel deformation prediction method contains the defects of a fixed model, a limited sample number, and it is easy to fall into underfitting and local overfitting. Therefore, the capacity of previous methods is limited by significant error, weak generalization, and poor intelligence. This paper proposes an adequate fitting prediction method for tunnel deformation based on the effective rank theory of the hidden layer nodes’ output matrix to analyze the surrounding rock and predict its deformation intelligently. Based on the traditional BPNN (back propagation neural network) algorithm, the number of hidden layer nodes is determined by the effective rank of the output matrix. Then, the approximation error and degree were adopted to reflect the approximation law of the BPNN to achieve the purpose of overfitting and underfitting control. An optimized BP neural network model for intelligently predicting tunnel deformation is constructed. Then, the optimized BPNN model is applied to a case study of a coastal tunnel in South China. Compared with the prediction method of LR (linear regression) and TS (time series), the results show that the prediction results of the optimized model are in good agreement with the measured values, with strong generalization ability and high intelligence. The proposed method is of guidance to other tunnels surrounding rock deformation prediction and engineering practice.

Reassessment of mobility parameters for Cantor High Entropy Alloys through an automated procedure

An automated assessment procedure is performed in order to establish a sophisticated kinetic data bank, introduced and modified by applying consequential iteration steps through the cross-validation method. The nonlinear curve-fitting of the end-member parameters is replaced by a simple linear fitting function via the logarithmic form of the Arrhenius equation. The applied modifications allow us to increase the precision of the method by decreasing the fitting errors. The input data employed here are the tracer diffusion coefficients in the well investigated high entropy alloy Co–Cr–Fe–Mn–Ni. The resulting parameters are in an acceptable agreement with the previously defined parameters in the literature while providing an efficient robust tool for kinetic data base development so that it enable an adequate prediction of diffusion transport.

Experimental and analytical studies on flexural behavior of composite bridge decks with orthotropic steel deck and ultra-high-performance concrete (UHPC) slab under negative moment

Ultra-high-performance concrete (UHPC) has been used to enhance the mechanical behavior of orthotropic steel bridge decks through steel-UHPC composite actions. This research develops a method to evaluate the flexural behavior of steel-UHPC composite sections via experimental and analytical investigations. Four composite deck specimens were fabricated and tested to study the mechanical behaviors under negative moment along the transverse and the longitudinal directions. A polyline model is proposed to link the strain, crack width of UHPC, and the bearing capacity of steel-UHPC composite decks, considering the crack behavior of UHPC with different fiber contents. The polyline model is used to derive engineer-friendly formulae for the flexural behavior of composite decks. Results showed that the proposed method provided adequate predictions for the flexural behavior of composite decks in the evolution of the crack width in UHPC with different fiber contents.

Pain predict genetics: protocol for a prospective observational study of clinical and genetic factors to predict the development of postoperative pain

IntroductionPostoperative pain remains a challenging medical condition impacting the quality of life of every patient. Although several predictive factors for postoperative pain have been identified, an adequate prediction of postoperative...

Air pollution prediction using LSTM deep learning and metaheuristics algorithms

Air pollution is a leading cause of health concerns and climate change, one of humanity's most dangerous problems. This problem has been exacerbated by an overabundance of automobiles, industrial output pollution, transportation fuel consumption, and energy generation. As a result, air pollution forecasting has become vital. As a result of the large amount and variety of data acquired by air pollution monitoring stations, air pollution forecasting has become a popular topic, particularly when applying deep learning models of long short-term memory (LSTM). The ability of these models to learn long-term dependencies in air pollution data sets them apart. However, LSTM models using many other statistical and machine learning approaches may not offer adequate prediction results due to noisy data and improper hyperparameter settings. As a result, to define the pollution levels for a group of contaminants, an ideal representation of the LSTM is required. To address the problem of identifying the best hyperparameters for the LSTM model, In this paper, we propose a model based on the Genetic Algorithm (GA) algorithm as well as the long short-term memory (LSTM) deep learning algorithm. The model aims to find the best hyperparameters for LSTM and the pollution level for the next day using four types of pollutants PM10, PM2.5, CO, and NOX. The proposed model modified by optimization algorithms shows more accurate results with less experience and more speed than machine learning models and LSTM models.

Multimodal Remote Sensing and Machine Learning for Precision Agriculture: A Review

This study proposed various machine learning and deep learning techniques to integrate and analyze varieties of data in precision agriculture systems. Agricultural systems have undergone a digital transformation, which has resulted in the evolution of many management components into artificially intelligent systems to better value the ever-increasing amounts of data. In the process of putting in place farming systems that are based on knowledge, several obstacles can be overcome using machine learning. The data obtained are transmitted to on-site storage, where extraction, loading, and transformation are performed. The data is preprocessed and transferred to the AWS (Amazon Web Services) cloud (Amazon S3 Bucket). The best model is deployed such that new data can be fit into the model to make adequate prediction or classification. Such a solution can be adapted by building an algorithm to simulate the AWS machine learning technique. A small-scale pilot project can be executed, and the output of the prediction or classification model can be displayed using a web-based software or mobile app.

Non-Linearity of Thermosetting Polymers' and GRPs' Thermal Expanding: Experimental Study and Modeling.

Thermal expanding is the important property that defines the stress–strain condition of GRP structures exploited under heating and having limited thermal resistance. So, the GRPs’ thermal expanding prediction is the actual requirement of such structures design. The experimental accurate dilatometric study resulted in the non-linearity of thermosetting polymers and plastics thermal expanding under heating. The polymers and plastics thermal expanding coefficient (CTE) is non-linearly increasing under heating before glassing temperature (Tg). Using the previous polymers and GRPs modelling experience and experimental dilatometric results, the non-linear adequate prediction models of their CTE were proposed and proved. The new compensative wave model of polymers’ CTE and multi-layer model of GRPs’ CTE were proposed and successfully tested. A prediction of the temperature dependences of the thermal expansion coefficients of various thermoset polymer binders and data on the reinforcement structure was performed based on the experimentally obtained temperature dependences of the CTEs of GRPs. The prediction was performed using the finite-element homogenization method in the Material Designer module of the academic version of the Ansys package. A satisfactory concurrence of the numerical results of the prognosis and the experiment for all considered cases is observed in the temperature range from 50 to 100 °C, after glass transition temperature best coincidence of numerical values of CTE is obtained for glass-reinforced plastics on epoxy resin, which were not subjected to thermal aging.

Machine Learning Approach to Predict In-Hospital Mortality in Patients Admitted for Peripheral Artery Disease in the United States.

Background Peripheral artery disease (PAD) affects >10million people in the United States. PAD is associated with poor outcomes, including premature death. Machine learning (ML) has been increasingly used on big data to predict clinical outcomes. This study aims to develop ML models to predict in-hospital mortality in patients hospitalized for PAD based on a national database. Methods and Results Inpatient hospitalization data were obtained from the 2016 to 2019 National Inpatient Sample. A total of 150 921 inpatients were identified with a primary diagnosis of PAD and PAD-related procedures using codes of the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) and International Classification of Diseases, Tenth Revision, Procedure Coding System (ICD-10-PCS). Four ML models, including logistic regression, random forest, light gradient boosting, and extreme gradient boosting models, were trained to predict the risk of in-hospital death based on a selection of variables, including patient characteristics, comorbidities, procedures, and hospital-related factors. In-hospital mortality occurred in 1.8% of patients. The performance of the 4 models was comparable, with the area under the receiver operating characteristic curve ranging from 0.83 to 0.85, sensitivity of 77% to 82%, and specificity of 72% to 75%. These results suggest adequate predictability for clinical decision-making. In all 4 models, the total number of diagnoses and procedures, age, endovascular revascularization procedure, congestive heart failure, diabetes, and diabetes with complications were critical predictors of in-hospital mortality. Conclusions This study demonstrates the feasibility of ML in predicting in-hospital mortality in patients with a primary PAD diagnosis. Findings highlight the potential of ML models in identifying high-risk patients for poor outcomes and guiding personalized intervention.

Meso-scale progressive damage simulation of textile composites using mesh superposition

The paper demonstrates that the mesh superposition (MSP) based finite element (FE) models of a textile composite unit cell can be used for adequate prediction of damage- and matrix plasticity-induced nonlinearity of the composite stress-strain response under tensile and shear loading. An elasto-plastic thermodynamically consistent damage model is implemented for matrix damage simulations. The intra-yarn damage initiation and propagation are simulated using Puck’s damage criterion and a continuous damage mechanics model. The numerical simulations are validated by a comparison with experimental results for tension and shear loading of a plain weave glass/epoxy laminate.