Performance Of Optimization Model Research Articles

Improving part-load performance of combined-cycle gas turbines by optimizing variable geometry control strategy for compressor and power turbine combined adjustment

The variable geometry methods currently used in combined-cycle gas turbines are compressor variable inlet guide vanes (VIGV) or power turbine variable area nozzles (VAN). On this basis, this study presents the optimal variable geometry control strategy for compressor and power turbine combined adjustment ([Formula: see text]) using the Differential Evolutionary Algorithm with the LM2500+ gas turbine. The aim is to further improve the part-load performance of the combined-cycle gas turbine. Firstly, a part-load performance prediction model for variable geometry gas turbines is established based on the component method. Subsequently, a variable geometry gas turbine part-load performance optimization model is developed by combining the Differential Evolution Algorithm. Finally, the optimum combination of stagger angles for the compressor inlet vane and power turbine nozzle is calculated at each part-load condition. Compared to the VIGV and VAN control strategies, the [Formula: see text] control strategy proposed in this paper shows a higher stability margin and better economy. The [Formula: see text] control strategy maintains a constant exhaust temperature within a part load range from 20% to 100% with the stability margin exceeding 14%. In comparison with the VAN control strategy, the fuel flow rate decreases by 1.152% at 45% relative load power and by 3.435% at 20.0% relative load power with the [Formula: see text] control strategy.

Diabetes Prediction Using Machine Learning

Diabetes, characterized by high blood sugar levels over time, poses challenges for accurate prediction due to limited labelled data and data quality issues like outliers and missing values. To address these challenges, our research introduces a robust prediction framework. This framework integrates techniques such as outlier rejection, filling missing values, standardizing data, selecting relevant features, and employing K-fold cross-validation. We utilize a range of machine learning algorithms including k-nearest Neighbour, Decision Trees, Random Forest, AdaBoost, Naive Bayes, XGBoost, and Multilayer Perceptron (MLP). Additionally, we propose a weighted ensembling technique to enhance prediction accuracy. The weights for ensembling are determined based on the performance of each classifier using the Area Under ROC Curve (AUC) metric. We optimize model performance through hyperparameter tuning using grid search. Our experiments are conducted using the Pima Indian Diabetes Dataset under uniform conditions. The ensembling classifier we propose achieves superior performance with a sensitivity of 0.789, specificity of 0.934, false omission rate of 0.092, diagnostic odds ratio of 66.234, and AUC of 0.950. This outperforms existing methods by 2.00% in terms of AUC. Our framework surpasses other approaches discussed in the literature and shows promise for improving diabetes prediction. We have released our source code for diabetes prediction to the public.

Structural- and DTI- MRI enable automated prediction of IDH Mutation Status in CNS WHO Grade 2–4 glioma patients: a deep Radiomics Approach

BackgroundThe role of isocitrate dehydrogenase (IDH) mutation status for glioma stratification and prognosis is established. While structural magnetic resonance image (MRI) is a promising biomarker, it may not be sufficient for non-invasive characterisation of IDH mutation status. We investigated the diagnostic value of combined diffusion tensor imaging (DTI) and structural MRI enhanced by a deep radiomics approach based on convolutional neural networks (CNNs) and support vector machine (SVM), to determine the IDH mutation status in Central Nervous System World Health Organization (CNS WHO) grade 2–4 gliomas.MethodsThis retrospective study analyzed the DTI-derived fractional anisotropy (FA) and mean diffusivity (MD) images and structural images including fluid attenuated inversion recovery (FLAIR), non-enhanced T1-, and T2-weighted images of 206 treatment-naïve gliomas, including 146 IDH mutant and 60 IDH-wildtype ones. The lesions were manually segmented by experienced neuroradiologists and the masks were applied to the FA and MD maps. Deep radiomics features were extracted from each subject by applying a pre-trained CNN and statistical description. An SVM classifier was applied to predict IDH status using imaging features in combination with demographic data.ResultsWe comparatively assessed the CNN-SVM classifier performance in predicting IDH mutation status using standalone and combined structural and DTI-based imaging features. Combined imaging features surpassed stand-alone modalities for the prediction of IDH mutation status [area under the curve (AUC) = 0.846; sensitivity = 0.925; and specificity = 0.567]. Importantly, optimal model performance was noted following the addition of demographic data (patients’ age) to structural and DTI imaging features [area under the curve (AUC) = 0.847; sensitivity = 0.911; and specificity = 0.617].ConclusionsImaging features derived from DTI-based FA and MD maps combined with structural MRI, have superior diagnostic value to that provided by standalone structural or DTI sequences. In combination with demographic information, this CNN-SVM model offers a further enhanced non-invasive prediction of IDH mutation status in gliomas.

Role of Education and Awareness Programs in Fostering Energy Conservation Behavior in Cities: Empowering Urban Sustainability Using Deep Learning Approach

This paper introduces a pioneering framework that emphasizes the critical role of education in promoting energy conservation behavior within urban environments. By synergizing deep learning methodologies with education-awareness formulations and algorithm, the framework aims to bolster energy conservation security in cities. A central aspect of this framework is the development of an advanced Intrusion Detection System (IDS) model, utilizing Restricted Boltzmann Machines (RBMs) tailored to enhance energy consumption security. Through six strategic integrations of educational components, the IDS not only identifies but also mitigates potential threats to energy systems with unprecedented efficiency. Additionally, the research introduces a novel evolutionary feature selection method based on Memetic Algorithms (MAs), which strategically identifies crucial features for RBM training, ensuring optimal model performance. The incorporation of Memetic Algorithms facilitates individual learning and problem-specific knowledge transmission, further enhancing the adaptive capabilities of the system. This multifaceted approach underscores the pivotal role of education in fostering urban sustainability, pushing the boundaries of energy conservation security. By proactively responding to emerging threats, the proposed framework contributes significantly to the overarching goal of sustainable and secure urban energy management.

The Phisherman’s Gambit

Phishing remains a pervasive threat in the digital landscape, posing significant risks to individuals and organizations alike. In response, this paper presents a sophisticated approach to phishing website detection utilizing machine learning methodologies. Through the amalgamation of diverse features extracted from URLs, domain attributes, and webpage content, a robust classification framework is constructed. Employing an array of supervised learning algorithms, including decision trees, support vector machines, and neural networks, our proposed system demonstrates its efficacy in accurately discerning phishing websites with notable precision and recall rates. Moreover, feature engineering and selection techniques are implemented to optimize model performance and computational efficiency. Experimental evaluations conducted on comprehensive datasets validate the effectiveness of the proposed approach in proactively identifying and mitigating phishing threats, thereby fortifying cybersecurity measures for users and organizations in the digital realm.

Multiple Disease Prediction System Using ML

Abstract: The increasing prevalence of diverse diseases presents a challenge to global healthcare systems, underscoring the need for innovative and efficient methods for early detection and preventive measures. This paper explores the application of machine learning algorithms in multiple disease prediction to enhance diagnostic accuracy and enable timely intervention. Leveraging diverse health-related data sources, including medical records and genomic information, comprehensive predictive models are developed. A multi-faceted machine learning approach integrates support vector machines, decision trees, neural networks, and ensemble learning methods to analyze complex data patterns. Feature selection and dimensionality reduction techniques optimize model performance and interpretability. The development of a predictive system involves essential steps such as data collection, preprocessing, and model training, followed by evaluation using metrics like accuracy and recall. Integration of Flask for web application development facilitates user interaction and prediction functionality. Deployment, testing, debugging, and ongoing maintenance ensure system efficiency and compliance with regulatory requirements for healthcare data security and privacy.

Streamlining event extraction with a simplified annotation framework.

Event extraction, grounded in semantic relationships, can serve as a simplified relation extraction. In this study, we propose an efficient open-domain event annotation framework tailored for subsequent information extraction, with a specific focus on its applicability to low-resource languages. The proposed event annotation method, which is based on event semantic elements, demonstrates substantial time-efficiency gains over traditional Universal Dependencies (UD) tagging. We show how language-specific pretraining outperforms multilingual counterparts in entity and relation extraction tasks and emphasize the importance of task- and language-specific fine-tuning for optimal model performance. Furthermore, we demonstrate the improvement of model performance upon integrating UD information during pre-training, achieving the F1 score of 71.16 and 60.43% for entity and relation extraction respectively. In addition, we showcase the usage of our extracted event graph for improving node classification in a retail banking domain. This work provides valuable guidance on improving information extraction and outlines a methodology for developing training datasets, particularly for low-resource languages.

A framework for electricity load forecasting based on attention mechanism time series depthwise separable convolutional neural network

Electricity load exhibits daily and weekly cyclical patterns as well as random characteristics. At present, prevailing deep learning models cannot learn electricity load cyclical and stochastic features adequately. This results in insufficient prediction accuracy and the scalability of current methods. To tackle these difficulties, this paper proposes a framework for electrical load prediction based on an Attention Mechanism Time Series Depthwise Separable Convolutional Neural Network (ELPF-ATDSCN). The framework starts by using the Maximum Information Coefficient for exogenous variable selection. It then incorporates a seasonal decomposition algorithm with manual feature engineering to extract the cyclical and stochastic features of the electrical load. Subsequently, the framework employs the ATDSCN to learn the cyclical and stochastic features of the electrical load. In addition, the Bayesian algorithm optimizes model hyperparameters for optimal model performance. Experimental results of point and interval load prediction on datasets from the US and Nordic power markets reveal that the ATDSCN model proposed in this paper enhances load prediction accuracy compared with other models. It can provide more reliable predictions for power system operation and dispatch.

Factors to improve odds of success following medial opening-wedge high tibial osteotomy: a machine learning analysis

BackgroundAlthough high tibial osteotomy (HTO) is an established treatment option for medial compartment osteoarthritis, predictive factors for HTO treatment success remain unclear. This study aimed to identify informative variables associated with HTO treatment success and to develop and internally validate machine learning algorithms to predict which patients will achieve HTO treatment success for medial compartmental osteoarthritis.MethodsThis study retrospectively reviewed patients who underwent medial opening-wedge HTO (MOWHTO) at our center between March 2010 and December 2015. The primary outcomes were a lack of conversion to total knee arthroplasty (TKA) and achievement of a minimal clinically important difference of improvement in the Knee Injury and Osteoarthritis Outcome Score (KOOS) at a minimum of five years postoperatively. Recursive feature selection was used to identify the combination of variables from an initial pool of 25 features that optimized model performance. Five machine learning algorithms (XGBoost, multilayer perception, support vector machine, elastic-net penalized logistic regression, and random forest) were trained using five-fold cross-validation three times and applied to an independent test set of patients. The performance of the model was evaluated by the area under the receiver operating characteristic curve (AUC).ResultsA total of 231 patients were included, and 200 patients (86.6%) achieved treatment success at the mean of 9 years of follow-up. A combination of seven variables optimized algorithm performance, and the following specific cutoffs increased the likelihood of MOWHTO treatment success: body mass index (BMI) ≤ 26.8 kg/m2, preoperative KOOS for pain ≤ 46.0, preoperative KOOS for quality of life ≤ 33.0, preoperative International Knee Documentation Committee score ≤ 42.0, preoperative Short-Form 36 questionnaire (SF-36) score > 42.25, three-month postoperative hip-knee-ankle angle > 1.0°, and three-month postoperative medial proximal tibial angle (MPTA) > 91.5° and ≤ 94.7°. The random forest model demonstrated the best performance (F1 score: 0.93; AUC: 0.81) and was transformed into an online application as an educational tool to demonstrate the capabilities of machine learning.ConclusionsThe random forest machine learning algorithm best predicted MOWHTO treatment success. Patients with a lower BMI, poor clinical status, slight valgus overcorrection, and postoperative MPTA < 94.7 more frequently achieved a greater likelihood of treatment success.Level of evidenceLevel III, retrospective cohort study.

Reliable fabric defect detection via Bayesian uncertainty modeling

Despite the demonstrated capability of deep learning models in detecting anomalies in textile images, their predictions in real-world applications tend to be overly confident, especially when faced with defect types not previously encountered in the training set or when dealing with low-quality annotations. This excessive confidence in predictions limits the practical application of deep learning methods in textile defect detection, as it fails to provide inspectors with reliable guidance on when to trust the model's predictions and when manual verification is necessary. To address this issue, this paper introduces a Bayesian fabric anomaly detection model that utilizes Variational Inference (VI) to apply Bayesian inference to the widely used U-Net architecture. During the inference phase, the model employs Monte Carlo sampling to perform multiple forward passes, generating three types of uncertainty estimations and per-pixel uncertainty maps, thus providing comprehensive evidence for decision-making. This method not only estimates the uncertainty of model predictions but also improves the F1 score by 2-4% over the baseline U-Net model in the frequency domain. This study proves the Bayesian approach boosts fabric anomaly detection and decision-making by optimizing model performance and reducing reliance on inaccurate predictions.

Efficient annotation reduction with active learning for computer vision-based Retail Product Recognition

AbstractThe retail industry encounters huge obstacles with computer vision (CV) technology due to frequent model retraining with changing products and time-consuming, costly data annotation. Previous research in this field has been primarily focused on optimizing model performance rather than minimizing annotation effort. Therefore, the main idea of this paper is to evaluate active learning as a method to minimize annotation effort in the retail industry. The MVTEC Densely Segmented Supermarket dataset is used to evaluate various active learning methods such as the Least Confident, Entropy and Cost-Effective Active Learning (CEAL) along with Mask R-CNN model. The results demonstrate that annotating only 20.83$$-$$ - 24.34% of the data achieves 95% of the full dataset’s performance. When training, out-of-sample data share similar characteristics, the Least Confident and CEAL methods reduce annotation requirements by 7.7$$-$$ - 15.7% while maintaining 95% and 97% of the full dataset’s performance. However, the Entropy method under-performs compared to the random selection baseline. Ultimately, none of the methods show a clear advantage when the data characteristics differ between training and out-of-sample data. Finally, the proposed active learning methods on an industry-specific retail dataset remarkably propels the development of highly efficient and cost-effective CV solutions meticulously tailored for the retail industry.

Optimization of chaotic light output in semiconductor laser systems based on multi-objective optimization algorithm.

Aiming at the weak performance of chaotic light output in semiconductor laser systems, the study designed a power control algorithm for semiconductor laser drive systems based on linear self-disturbance rejection control. Then the optimization parameters and scope were determined, and multi-objective optimization and direction preference algorithms were introduced. A chaotic optical performance optimization model based on improved multi-objective genetic algorithm was constructed using adaptive functions as evaluation indicators. These results confirmed that the larger the bandwidth of the controller, the faster the response speed of the resonant converter, but the stability was poor. When the input voltage underwent a sudden change, the current ripple coefficient of the PID algorithm was 0.55%. The linear active disturbance rejection control algorithm could ensure that the voltage and current maintained at the set values, and the output current of the algorithm was more stable when the load underwent sudden changes. The directional preference algorithm could further provide more valuable solutions on the basis of adaptive genetic algorithms. When the peak value of the autocorrelation function was equal to 0.2, the delay characteristics of chaotic light were effectively suppressed, having strong signal bandwidth and complexity. In summary, the constructed model has good application effects in optimizing chaotic optical performance and has certain positive significance for communication security.

Towards efficient similarity embedded temporal Transformers via extended timeframe analysis

AbstractPrice prediction remains a crucial aspect of financial market research as it forms the basis for various trading strategies and portfolio management techniques. However, traditional models such as ARIMA are not effective for multi-horizon forecasting, and current deep learning approaches do not take into account the conditional heteroscedasticity of financial market time series. In this work, we introduce the similarity embedded temporal Transformer (SeTT) algorithms, which extend the state-of-the-art temporal Transformer architecture. These algorithms utilise historical trends in financial time series, as well as statistical principles, to enhance forecasting performance. We conducted a thorough analysis of various hyperparameters including learning rate, local window size, and the choice of similarity function in this extension of the study in a bid to get optimal model performance. We also experimented over an extended timeframe, which allowed us to more accurately assess the performance of the models in different market conditions and across different lengths of time. Overall, our results show that SeTT provides improved performance for financial market prediction, as it outperforms both classical financial models and state-of-the-art deep learning methods, across volatile and non-volatile extrapolation periods, with varying effects of historical volatility on the extrapolation. Despite the availability of a substantial amount of data spanning up to 13 years, optimal results were primarily attained through a historical window of 1–3 years for the extrapolation period under examination.

Development of Multi Modal Image Fusion Techniques in Night Vision Technology Using Deep Learning

This study is aimed at exploring new development in nighttime image processing based on nighttime photography/night vision. We use various learning methodologies based on neural networks to develop multisensor data processing techniques. The ability of Convolutional Neural Networks (CNN) to enhance image quality and clarity under conditions of low light or at night is the focus of our research. To do this A large database has been created, containing 3200 images both taken from the internet as well as our own. This data is preprocessed, normalized and features expanded using some common methods of data augmentation in order to optimize model performance. We compared our proposed CNN model with existing literature through a series of experiments. Performance was measured using Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Feature Similarity Index (FSIM). The Results are now in. They showed that our model can improve various metrics with higher scores. And they show the real-time monitoring capabilities in urban streets for our model. This research brings multisensor image fusion based on night-adjacent techniques forward from night vision technology and provides insight into practical and potential future directions for using deep learning techniques to improve safety and security in many different environments.

Machine Learning for Early Diabetes Detection and Diagnosis

In this work, a machine learning (ML)-based e-diagnostic system is suggested specifically for the detection of gestational diabetes mellitus (GDM). Reviewing recent GDM data and outlining the intimate connection between GDM and prediabetic conditions, as well as the potential for future declines in insulin resistance and the emergence of overt Type 2 diabetes, were our goals. The present study explores the application of the K-nearest neighbors (KNN) algorithm to project diabetes diagnosis on the widely-used Pima Indians Diabetes database. The KNN algorithm, a non-parametric, instance-based learning method, was employed to classify individuals as either diabetic or non-diabetic, our objectives were to evaluate the algorithm’s ability to make accurate predictions and explore factors influencing its performance. The study commenced with data preprocessing, including handling missing values, feature scaling, and data splitting into training and testing sets. The KNN classifier was trained and tested using these best-fit parameters. The results of this study revealed a model with an accuracy of approximately 0.76 in predicting diabetes diagnosis. This study looked at the various machine-learning approaches for diabetes patient classification, including recall, accuracy, precision, and F1-score. The study discusses the significance of hyperparameter tuning, data preprocessing, and imbalanced data handling in achieving optimal KNN model performance. Lastly, this study shows how the KNN algorithm may be used to project diabetes using the Pima Indians Diabetes Database. The findings suggest that KNN can serve as a viable tool in the early detection of diabetes, paving the way for more extensive applications in healthcare and predictive modelling.

Improving parameterization of an evapotranspiration estimation model with eddy covariance measurements for a regional irrigation scheduling program

Actual evapotranspiration (ETa) is an essential variable in linking energy cycles, carbon, and water, yet challenging to measure. Inputs uncertainty and deficiencies in the key elements of hydrologic models are fundamental challenges for optimizing model performance. Furthermore, the performance of land surface model-based ETa, reanalysis, and remote sensing products varies with spatiotemporal scales. Here, we evaluate sources of bias in the regional Wisconsin Irrigation and Scheduling Program (WISP) model and develop a correction using eddy covariance (EC) observations. ETa, observations were made for five years (2018–2022) using EC systems in agricultural fields in Wisconsin. WISP model ETa bias was linked to underestimation of net longwave radiation (LWnet) that was traced to incorrect specification of effective clear sky atmospheric emissivity (εa,clr). Applying a correction to the εa,clr led to reduced WISP model percent bias (pbias) and error for both LWnet and ETa. The calibrated model more accurately represented observed ETa. The results indicate that explicit treatment of the LWnet balance decreases the uncertainty of model parameters and improves the WISP model performance at independent sites. Applying this improved model parameterization reduced the bias of LWnet radiation from 62.8% to -6.2%, which improved the Nash-Sutcliffe Efficiency (NSE) from -0.08 to 0.52 for ETa on training sites. Additionally, overall pbias was significantly reduced (p = 0.035) for validation sites after WISP correction. Hence, WISP performance improved for different crop types when optimal regional parameters were used, confirming the physical parameters' reliability. Our results highlight that model development should focus on energy balance parameterizations to improve ET simulation and the accuracy of hydrologic and climatic simulations for understanding critical processes underlying hydrologic and climatic variability and change over land.

Multi-objective optimization and evaluation of hybrid combined cooling, heating and power system considering thermal energy storage

Introducing thermal energy storage (TES) and solar energy effectively reduces fossil fuel consumption and greenhouse gas emissions in combined cooling, heating, and power (CCHP) systems. This study establishes a two-layer optimization framework to obtain the optimal configuration of the CCHP system coupling solar and thermal energy storage. The outer layer uses component sizes as decision variables to optimize the system's annual economic economy, annual carbon dioxide emissions, and annual primary energy consumption; the inner layer uses equipment output as decision variables to optimize the operation and maintenance costs. Firstly, the representative typical scenes were obtained using the k-means algorithm to reduce the computational complexity. Then, the multi-objective chaos optimization (MOCGO) algorithm is combined with mixed-integer linear programming (MILP) to obtain the established optimization model's configuration scheme and optimal operating strategy. TOPSIS is to rank the best Pareto solutions. Finally, a large hotel is a case study. The optimization results indicate that the proposed two-layer optimization model can achieve the optimal configuration scheme and scheduling strategy for hybrid CCHP. Compared to the rule-based multi-objective one-layer optimization model, the proposed two-layer optimization model can achieve maximum improvements of 30.63 %, 37.94 %, and 38.16 % in economic, energy, and environmental performance. Compared with the rule-based single-objective one-layer optimization model, the proposed two-layer optimization model can achieve maximum improvements of 34.95 %, 35.81 %, and 37.55 % in economic, energy, and environmental performance, respectively. In addition, TES can substantially increase the optimal performance of the two-layer optimization model.

Anomaly‐based Alzheimer's disease detection using entropy‐based probability Positron Emission Tomography images

AbstractDeep neural networks trained on labeled medical data face major challenges owing to the economic costs of data acquisition through expensive medical imaging devices, expert labor for data annotation, and large datasets to achieve optimal model performance. The heterogeneity of diseases, such as Alzheimer's disease, further complicates deep learning because the test cases may substantially differ from the training data, possibly increasing the rate of false positives. We propose a reconstruction‐based self‐supervised anomaly detection model to overcome these challenges. It has a dual‐subnetwork encoder that enhances feature encoding augmented by skip connections to the decoder for improving the gradient flow. The novel encoder captures local and global features to improve image reconstruction. In addition, we introduce an entropy‐based image conversion method. Extensive evaluations show that the proposed model outperforms benchmark models in anomaly detection and classification using an encoder. The supervised and unsupervised models show improved performances when trained with data preprocessed using the proposed image conversion method.

Harnessing the power of radiomics and deep learning for improved breast cancer diagnosis with multiparametric breast mammography

Breast cancer, with its high mortality, faces diagnostic challenges due to variability in mammography quality and breast densities, leading to inconsistencies in radiological interpretations. Computer-aided diagnostic (CAD) systems, while helpful, struggle with accurately interpreting lesion characteristics such as morphology, density, and size. To address this, our study developed advanced deep-learning algorithms to improve the detection, localization, risk assessment, and classification of breast lesions, aiming to reduce false positives on human intervention and tackle slow convergence rates. Key innovations of the approach include preprocessing techniques, advanced filtering, and data augmentation strategies to optimize model performance, mitigating over- and under-fitting concerns. A significant development is the Chaotic Leader Selective Filler Swarm Optimization (cLSFSO) method, which effectively detects breast-dense lesions by extracting textural and statistical features. Additionally, the study adapted deep learning models like modified VGGNet and SE-ResNet152 through transfer learning, significantly enhancing their capability to distinguish between normal and suspicious mammography regions. The study also introduces hybrid deep neural network-based approaches, including CNN+LSTM and CNN+SVM, for diagnosing and grading cancerous polyps from the pre-segmented ROIs. Besides, the transfer learning paradigm is employed to boost the efficacy in classifying breast masses and reducing computing time by modifying the final layer of the proposed pre-trained models. The integration of Grad-CAM techniques further refines our analysis, leading to more accurate assessments and improved diagnosis of breast anomalies. Evaluated using benchmark and private datasets, our algorithms demonstrated a sensitivity of 0.99 and an overall AUC of 0.99, indicating significant improvements in mammogram analysis. These advancements aid radiologists, potentially improving patient outcomes and contributing to medical imaging and AI in healthcare.

Bayesian analysis of optimal burn‐in policy for heterogeneous components with minimal repair

AbstractThis study develops a Bayesian method to analyze optimal burn‐in policy for heterogeneous components with minimal repair. We use the non‐homogeneous Poisson processes with different power law intensity functions to model the minimal repair processes for heterogeneous components. Since the component from the weak subpopulation may fail more frequently, the total number of minimal repairs can be used for establishing a screening rule. By screening, we can decide whether the component after burn‐in should be eliminated or not. Considering the uncertainty of model parameters, we generate a minimally repaired data set of components by simulation. Then, we obtain the model parameters estimators by using Bayesian method combined with the existing data. We further analyze and get the optimal burn‐in settings under cost and performance optimization model. A numerical example is used to verify the effectiveness of the Bayesian method.