Evaluation Metrics Research Articles

Research on coal rock parameter calibration based on discrete element method

To ensure that the discrete element model of the coal wall accurately reflects the actual cutting process of coal and rock during virtual prototype simulation of mining equipment, this study aims to expedite the development of intelligent mining machinery. Using coal samples from the 4602 working face of Yangcun Coal Mine, operated by Yanzhou Coal Mining Group, we conducted coal rock packing experiments and uniaxial compression tests to obtain the packing angle and compressive strength of the coal samples. Based on the experimental results, we designed Plackett–Burman experiments (PB experiments), steepest ascent experiments, and Box-Behnken experiments to study the influence of particle contact mechanics parameters and bonding mechanics parameters on the packing angle and compressive strength, using the packing angle and compressive strength as response variables. Our objective is to minimize the relative error between the simulated packing angle and the measured packing angle. We solved and optimized the parameter calibration model and conducted simulation calibration experiments based on the optimization results. A comparative analysis with the actual test results revealed that the maximum relative errors between the simulated and measured values for the packing angle and compressive strength were only 2.9% and 5.0%, respectively. Additionally, a discrete element model of a typical working face coal wall was established based on the parameters obtained from this calibration method. A bidirectional coupling model of the cutting process between the coal and rock was created using EDEM-RecurDyn to simulate the rigid-flexible coupling of the coal cutter. An experimental coal wall model was constructed based on similarity theory, and both simulation and physical experiments were conducted. The evaluation metrics for comparison were the time-domain and frequency-domain characteristics of the drum’s vibration signals. The maximum relative error for the time-domain signal characteristics between the two experimental setups was only 4.19%, while the maximum relative error for the frequency-domain signal characteristics was 3.75%. This validates the feasibility of the proposed calibration method for the discrete element coal wall model and the accuracy of the calibration results. Furthermore, it demonstrates that the constructed discrete element coal wall accurately represents the actual coal and rock properties. The virtual simulation based on this model effectively replicates the interaction process between mining machinery and coal, providing a safe, efficient, and low-cost technological approach for performance analysis of mining equipment and intelligent control of supporting devices.

An intelligent monitoring approach for urban natural gas pipeline leak using semi-supervised learning generative adversarial networks

Traditional gas pipeline leak monitoring methods are subjected to the long response times and high false alarm rates. Deep learning can enhance the accuracy and real-time performance of pipeline leak monitoring. This paper develops an intelligent monitoring approach for urban gas pipeline leaks based on a semi-supervised learning Generative Adversarial Network (SGAN). First, the Isolation Forest algorithm is used to classify anomalies in the collected process parameter data of urban natural gas pipelines. One-Hot Encoding is used to label a small amount of sample data of pipeline leak. Second, both the labeled and unlabeled data are input into SGAN model for semi-supervised learning and classification to monitor the state of urban gas pipeline leak. The methodology addresses the imbalance between pipeline leak status data and normal data. The comparison with GAN and MLP shows that the methodology reaches the highest values in all evaluation metrics (precision = 94.1%, accuracy = 95.63%, recall = 93.93%, F1 score = 94.32%). The superior performance and accuracy make it more effective for urban natural gas pipeline leak monitoring.

Detecting Water Stress in Winter Wheat Based on Multifeature Fusion from UAV Remote Sensing and Stacking Ensemble Learning Method

The integration of remote sensing technology and machine learning algorithms represents a new research direction for the rapid and large-scale detection of water stress in modern agricultural crops. However, in solving practical agricultural problems, single machine learning algorithms cannot fully explore the potential information within the data, lacking stability and accuracy. Stacking ensemble learning (SEL) can combine the advantages of multiple single machine learning algorithms to construct more stable predictive models. In this study, threshold values of stomatal conductance (gs) under different soil water stress indices (SWSIs) were proposed to assist managers in irrigation scheduling. In the present study, six irrigation treatments were established for winter wheat to simulate various soil moisture supply conditions. During the critical growth stages, gs was measured and the SWSI was calculated. A spectral camera mounted on an unmanned aerial vehicle (UAV) captured reflectance images in five bands, from which vegetation indices and texture information were extracted. The results indicated that gs at different growth stages of winter wheat was sensitive to soil moisture supply conditions. The correlation between the gs value and SWSI value was high (R2 > 0.79). Therefore, the gs value threshold can reflect the current soil water stress level. Compared with individual machine learning models, the SEL model exhibited higher prediction accuracy, with R2 increasing by 6.67–17.14%. Using a reserved test set, the SEL model demonstrated excellent performance in various evaluation metrics across different growth stages (R2: 0.69–0.87, RMSE: 0.04–0.08 mol m−2 s−1; NRMSE: 12.3–23.6%, MAE: 0.03–0.06 mol m−2 s−1) and exhibited excellent stability and accuracy. This research can play a significant role in achieving large-scale monitoring of crop growth status through UAV, enabling the real-time capture of winter wheat water deficit changes, and providing technical support for precision irrigation.

Semantic segmentation of optical satellite images for the illegal construction detection using transfer learning

As urban populations grow, the demand for new infrastructure intensifies. However, some builders often neglect to obtain necessary permits, resulting in unstable structures that endanger public safety and disrupt city planning. Illegal construction poses significant challenges to economic development and social harmony. To tackle this issue, there is an urgent need for an automated illegal building monitoring system that can accurately identify and notify authorities about unlawful constructions. Current methods are often inefficient due to limited data access, inadequate analysis, and lack of automation.Recent technological advancements suggest that using satellite imagery alongside semantic segmentation is a highly effective approach for detecting illegal buildings. This process not only ensures the safety and compliance of structures but also enhances overall regulatory enforcement. This research distinguishes itself from previous studies by primarily focusing on illegal construction. It employs a diverse range of machine learning models, including the U-Shaped Network and Visual Geometry Group, and incorporates customized evaluation metrics that are not typically found in earlier research. Additionally, it utilizes real-world data, enhancing its practical relevance, and features IoT integration for real-time monitoring capabilities.We chose U-Net as our architecture due to its symmetrical design, which facilitates effective feature extraction. Its skip connections play a crucial role in preserving important features during the segmentation process. U-Net is known for its high accuracy and adaptability across various domains, making it versatile for our needs. Additionally, it performs well even with limited data, demonstrating robustness in diverse conditions. Furthermore, its compatibility with transfer learning enhances its applicability. For all these reasons, U-Net was selected as our model.Our experiments revealed that the Unet_mini model outperformed all others, demonstrating its effectiveness in addressing illegal construction challenges in urban areas. Qualitative results further illustrate the model's robustness, showing a significant reduction in false positives and improved accuracy in identifying illegal structures. This comprehensive approach not only improves detection rates but also contributes to more efficient regulatory enforcement, thereby promoting safer urban environments.This research distinguishes itself from previous studies by primarily focusing on illegal construction detection rather than general segmentation tasks. It employs a diverse range of machine learning models, including the U-Shaped Network and Visual Geometry Group, and integrates customized evaluation metrics that are not typically found in earlier research. Furthermore, it utilizes real-world data, enhancing its practical relevance, and incorporates IoT integration for real-time monitoring capabilities, setting it apart from existing studies.

An Implementation of XGBoost and Random Forest Algorithm to Estimate Effective Porosity and Permeability on Well log data at Fajar Field, South Sumatera Basin, Indonesia

<p class="Keywords">New approaches and methodologies have been developed to petrophysical analysis from well logs data using machine learning. Through this method, a machine learning algorithm is applied to predict the accuracy of the model on effective porosity (∅e) and permeability (K) which implemented using Random Forest and Xtreme Gradient Boosting (XGBoost) algorithm. The dataset used is obtained from well logs data that have been calculated petrophysical analysis. This study proposes the algorithm which is known to be effective in providing accurate predictions in a short time in estimating effective porosity and permeability. The results of the prediction model is optimized by GridSearchCV (GS), validated by the k-fold cross-validation, and evaluated using R2 score and Root Mean Square Error (RMSE). Model is applied to 5 research wells in Fajar Field of south Sumatra Basin, Indonesia with 4 variations of well training and well testing data split. The best evaluation results obtained with evaluation metrics were up to 0.90 (R2 score) and 0.01 (RMSE) for effective porosity and permeability by Random Forest, while evaluation metrics are 0.90 (R2 score) and under 0.68 (RMSE) for effective porosity and permeability by XGBoost. There is no decrease in accuracy until the last variation so that it can be concluded that these algorithm models can effectively estimate reservoir porosity and permeability in the field and contributed an alternative for the problem of many incomplete and dissimilar well logs data.</p>

Forensic bone age assessment of hand and wrist joint MRI images in Chinese han male adolescents based on deep convolutional neural networks.

In Chinese criminal law, the ages of 12, 14, 16, and 18 years old play a significant role in the determination of criminal responsibility. In this study, we developed an epiphyseal grading system based on magnetic resonance image (MRI) of the hand and wrist for the Chinese Han population and explored the feasibility of employing deep learning techniques for bone age assessment based on MRI of the hand and wrist. This study selected 282 Chinese Han Chinese males aged 6.0-21.0 years old. In the course of our study, we proposed a novel deep learning model for extracting and enhancing MRI hand and wrist bone features to enhance the prediction of target MRI hand and wrist bone age and achieve precise classification of the target MRI and regression of bone age. The evaluation metric for the classification model including precision, specificity, sensitivity, and accuracy, while the evaluation metrics chosen for the regression model are MAE. The epiphyseal grading was used as a supervised method, which effectively solved the problem of unbalanced sample distribution, and the two experts showed strong consistency in the epiphyseal plate grading process. In the classification results, the accuracy in distinguishing between adults and minors was 91.1%, and the lowest accuracy in the three minor classifications (12, 14, and 16 years of age) was 94.6%, 91.1% and 96.4%, respectively. The MAE of the regression results was 1.24 years. In conclusion, the deep learning model proposed enabled the age assessment of hand and wrist bones based on MRI.

Anatomical prior-based vertebral landmark detection for spinal disorder diagnosis

As one of fundamental ways to interpret spine images, detection of vertebral landmarks is an informative prerequisite for further diagnosis and management of spine disorders such as scoliosis and fractures. Most existing machine learning-based methods for automatic vertebral landmark detection suffer from overlapping landmarks or abnormally long distances between nearby landmarks against anatomical priors, and thus lack sufficient reliability and interpretability. To tackle the problem, this paper systematically utilizes anatomical prior knowledge in vertebral landmark detection. We explicitly formulate anatomical priors of the spine, related to distances among vertebrae and spatial order within the spine, and integrate these geometrical constraints within training loss, inference procedure, and evaluation metrics. First, we introduce an anatomy-constraint loss to regularize the training process with the aforementioned contextual priors explicitly. Second, we propose a simple-yet-effective anatomy-aided inference procedure by employing sequential prediction rather than a parallel counterpart. Third, we provide novel anatomy-related metrics to quantitatively evaluate to which extent landmark predictions follow the anatomical priors, as is not reflected within the widely-used landmark localization error metric. We employ the localization framework on 1410 anterior-posterior radiographic images. Compared with competitive baseline models, we achieve superior landmark localization accuracy and comparable Cobb angle estimation for scoliosis assessment. Ablation studies demonstrate the effectiveness of designed components on the decrease of localization error and improvement of anatomical plausibility. Additionally, we exhibit effective generalization performance by transferring our detection method onto sagittal 2-D slices of CT scans and boost the performance of downstream compression fracture classification at vertebra-level.

Validation of different automated segmentation models for target volume contouring in postoperative radiotherapy for breast cancer and regional nodal irradiation

IntroductionTarget volume delineation is routinely performed in postoperative radiotherapy (RT) for breast cancer patients, but it is a time-consuming process. The aim of the present study was to validate the quality, clinical usability and institutional-specific implementation of different auto-segmentation tools into clinical routine. MethodsThree different commercially available, artificial intelligence-, ESTRO-guideline-based segmentation models (M1-3) were applied to fifty consecutive reference patients who received postoperative local RT including regional nodal irradiation for breast cancer for the delineation of clinical target volumes: the residual breast, implant or chestwall, axilla levels 1 and 2, the infra- and supraclavicular regions, the interpectoral and internal mammary nodes. Objective evaluation metrics of the created structures were conducted with the Dice similarity index (DICE) and the Hausdorff distance, and a manual evaluation of usability. ResultsThe resulting geometries of the segmentation models were compared to the reference volumes for each patient and required no or only minor corrections in 72 % (M1), 64 % (M2) and 78 % (M3) of the cases. The median DICE and Hausdorff values for the resulting planning target volumes were 0.87–0.88 and 2.96–3.55, respectively. Clinical usability was significantly correlated with the DICE index, with calculated cut-off values used to define no or minor adjustments of 0.82–0.86. Right or left sided target and breathing method (deep inspiration breath hold vs. free breathing) did not impact the quality of the resulting structures. ConclusionArtificial intelligence-based auto-segmentation programs showed high-quality accuracy and provided standardization and efficient support for guideline-based target volume contouring as a precondition for fully automated workflows in radiotherapy treatment planning.

Image Diversity Evaluation Metrics Correlated with Human Subjectivity and Prediction of Image Diversity in Text-to-image Synthesis

Image Diversity Evaluation Metrics Correlated with Human Subjectivity and Prediction of Image Diversity in Text-to-image Synthesis

MDDU-Net: A multi-scale dense connectivity hybrid dilated convolutional U-Net for segmentation in prostate ultrasound images

Ultrasound is one of the most commonly used imaging tools in prostate biopsy and brachytherapy. However, due to issues such as blurred image boundaries, similar intensity distributions, and noise, accurate segmentation of the prostate gland has been challenged. We propose a multi-scale dense connectivity hybrid dilated convolutional U-Net (MDDU-Net) network model segmentation of 2D prostate ultrasound images to improve the segmentation accuracy. Our proposed hybrid dilated convolution (HDC) module consists of standard and dilated convolution, which can expand the receptive field of the network, focus on global information, capture more features, and improve the learning ability of the model. Introducing dense connectivity (Dense) maximizes the flow of information between layers, retains more details, and helps mitigate gradient vanishing. The output features of each convolution layer in the bottleneck and decoder are fused using a multi-scale fusion block (MSFB) to obtain a feature map with richer semantic and detailed information. A dense skip connection provides more raw information to the decoder. To verify the superiority of our proposed method, we performed a 5-fold cross-validation, quantitatively analyzed it using five commonly used medical image evaluation metrics, and compared it with several state-of-the-art segmentation methods. The results show that our method performs better in prostate ultrasound image segmentation and can achieve more accurate prostate segmentation.

New metrics for measuring 2D uniformity in stirring system based on reconstruction of the particle trajectory

In industrial mixing systems, the evaluation method is a crucial indicator for measuring mixing uniformity. This study proposes an assessment method that combines a method for positioning using dual cameras and a point pattern density fluctuation (PD) method based on disordered hyperuniformity. This study employs a method for positioning using dual cameras to investigate the motion trajectories of tracer balls and achieve a 3D reconstruction process. The reconstructed 3D systems are cross-sectioned through on the axis of the stirring paddle to form a 2D planar diagram of the tracer ball trajectories. The PD method is used to evaluate the mixing effectiveness of the 2D trajectory points, establishing evaluation metrics for the 2D mixing system. To ensure the accuracy of the PD method, a feasibility analysis was conducted using the conductometry. A relationship model between λ and mixing time was established, with a Pearson correlation coefficient (r) of -0.9867, indicating a strong negative linear correlation. The relationship between the planes and whole 3D space was analyzed in terms of the whole and the discrete perspective, as well as the existing methods were compared with other approaches. This study not only deepens the theoretical understanding of evaluating mixing uniformity but also provides a crucial foundation and theoretical basis for process optimization of stirring systems and design of stirred reactors. It offers valuable references and insights for engineering design and technological improvements in related fields.

SIPSC-Kac: Integrating swarm intelligence and protein spatial characteristics for enhanced lysine acetylation site identification

Elucidation of post-translational modifications (PTMs), such as lysine acetylation (Kac), is crucial for understanding protein function and regulation. Although traditional experimental methods for identifying Kac sites are accurate, they are time-consuming and costly, leading to incomplete acetylome mapping. Computational approaches, particularly those incorporating machine learning, offer a rapid alternative, but face challenges owing to dataset imbalance, limited feature space, and the need for more effective feature-selection algorithms. To address these challenges, we present SIPSC-Kac, a novel computational method that integrates swarm intelligence algorithms with protein spatial characteristics to enhance the prediction of Kac sites. We used the AlphaFold system for spatial feature extraction and employed swarm intelligence for optimal feature selection, outperforming existing methods in terms of accuracy and computational efficiency. SIPSC-Kac demonstrated superior performance across multiple bacterial species, which was validated by its high performance in evaluation metrics. Our web server provides researchers with a user-friendly platform for Kac site prediction, thereby contributing to the advancement of bioinformatics and proteomic research. The SIPSC-Kac code and web server are accessible, thereby promoting broad applications in the scientific community.

Ensemble deep learning and anomaly detection framework for automatic audio classification: Insights into deer vocalizations

Audio recordings have emerged as a pivotal tool in field observations, enriching environmental monitoring in both the spatial and temporal dimensions. However, the richness and complexity of these recordings pose significant challenges, primarily when extracting specific sound clips from long recordings owing to the presence of ambient noise and other irrelevant sounds. Traditional methods, such as manual extraction or a sliding window over audio segments, hinder practical bioacoustic applications. Therefore, we propose a framework that begins with a robust segmentation method for extracting sound clips that potentially contain deer vocalizations. This segmentation method relies on acoustic anomaly detection and can markedly improve computational efficiency, facilitating deployment in environments with limited resources. Subsequently, the isolated clips were classified into deer and non-deer categories using machine learning models. Our investigation assessed three state-of-the-art deep learning models, ResNet50, MobileNetV2, and EfficientNet-B2, considering various hyperparameter configurations to optimize the performance. We utilized 3842 clips from two sites, Oze National Park and Taki, for training and testing. The outcomes demonstrated that all models exhibited comparable performances, with median accuracies of 98.3 % and 92.9 % during the validation and testing stages, respectively. However, no single model outperformed the others across all the evaluation metrics. For instance, ResNet50 in different configurations led to the best accuracy, F1 score, precision, and specificity, whereas MobileNetV2 had the best recall. Therefore, we adopted a consensus-based ensemble scoring system in which an audio clip was classified as a deer call when at least two of three models concurred in their classification to enhance the reliability of our classifications. Our findings demonstrated that the Ensemble approach significantly enhanced the classification performance, achieving an accuracy of 99.2 % in the test stage. The proposed approach was successfully deployed during the deer rutting seasons in Oze and Taki in 2019 and 2021, respectively. We gained invaluable insights into deer behavior by analyzing deer calls' frequency, timing, and duration. Additionally, the spatial distribution of deer calls in Taki enabled us to detect a breach in the city's protective fencing and an association between the spatial patterns of deer calls and crop damage in the two fields. We aimed to draw a comprehensive picture of deer activity, which has significant implications for both conservation efforts and understanding animal behavior in various habitats. The insights gathered from this research contribute to the scientific understanding of deer behavior and serve as a foundation for future studies and conservation initiatives. By incorporating advanced machine learning models into environmental monitoring, we have paved the way for more data-driven approaches in wildlife research.

Assessing the impact of the National Clean Air Programme in Uttar Pradesh's non-attainment cities: a prophet model time series analysis

BackgroundUttar Pradesh, India’s largest state, faces critical pollution levels, necessitating urgent action. The National Clean Air Programme (NCAP) targets a 40% reduction in particulate pollution by 2026. This study assesses the NCAP’s impact on 15 non-attainment cities in Uttar Pradesh using the Prophet forecasting model. MethodsMonthly data from 2016 to 2023 on AQI and PM10 concentrations were sourced from the Uttar Pradesh Pollution Control Board. Significant changes in mean AQI and PM10 levels from 2017 to 2023 were evaluated using the Friedman Test. Prophet models forecasted PM10 concentrations for 2025-26, with relative percentage changes calculated and model evaluation metrics assessed. FindingsMost cities exhibited unhealthy air quality. Jhansi had the lowest AQI (72·73) in 2023, classified as 'moderate' by WHO standards. Gorakhpur consistently showed 'poor' AQI levels, peaking at 249·31 in 2019. Western Uttar Pradesh cities such as Ghaziabad, Noida, and Moradabad had significant pollution burdens. Predictions showed Bareilly with over a 70% reduction in PM10 levels, Raebareli 58%, Moradabad 55%, Ghaziabad 48%, Agra around 41%, and Varanasi 40%, meeting NCAP targets. However, Gorakhpur and Prayagraj predicted increases in PM10 levels by 50% and 32%, respectively. Moradabad’s model showed the best performance with an R2 of 0·81, MAE of 17·27 μg/m3, and MAPE of 0·10. InterpretationForecasting PM10 concentrations in Uttar Pradesh's non-attainment cities offers policymakers substantial evidence to enhance current efforts. While existing measures are in place, our findings suggest that intensified provisions may be necessary for cities predicted to fall short of meeting program targets. The Prophet model's forecasts can pinpoint these at-risk areas, allowing for targeted interventions and regional adjustments to strategies. This approach will help promote sustainable development customized to each city's specific needs. FundingNo funding was issued for this research.

Multiresolution Analysis of HRRR Meteorological Parameters and GOES-R AOD for Hourly PM2.5 Prediction.

High-resolution, comprehensive exposure data are crucial for accurately estimating the human health impact of PM2.5. In recent years, satellite remote sensing data have been increasingly utilized in PM2.5 models to overcome the limited spatial coverage of ground monitoring stations. However, data gaps in satellite-retrieved parameters such as aerosol optical depth (AOD), the sparsity of regulatory air quality monitors for model training, and nonlinear relationships between PM2.5 and meteorological conditions can affect model performance and cause data gaps in most existing PM2.5 models. In this study, spatial gaps in AOD obtained from Geostationary Operational Environmental Satellite-16 are filled using Goddard Earth Observing System Composition Forecasting AOD estimations. Furthermore, to improve model performance, meteorological predictors such as temperature from the High-Resolution Rapid Refresh model are preprocessed using Daubechies wavelet to extract low and high frequency components. The spatially gap-filled AOD, along with meteorological data, are ingested into a machine learning model to predict hourly PM2.5 at a 1 km spatial resolution in California. The model evaluation metrics (OOB (out-of-bag) R2 = 0.86 and RMSE (root-mean-square error) = 9.27 μg/m3 and 10-fold spatial cross-validation R2 = 0.82 and RMSE = 9.82 μg/m3) demonstrate the model's reliability in predicting ambient PM2.5, especially for states like California that experience frequent episodes of wildfires.

Enhancing early detection of cognitive decline in the elderly: a comparative study utilizing large language models in clinical notes

BackgroundLarge language models (LLMs) have shown promising performance in various healthcare domains, but their effectiveness in identifying specific clinical conditions in real medical records is less explored. This study evaluates LLMs for detecting signs of cognitive decline in real electronic health record (EHR) clinical notes, comparing their error profiles with traditional models. The insights gained will inform strategies for performance enhancement. MethodsThis study, conducted at Mass General Brigham in Boston, MA, analysed clinical notes from the four years prior to a 2019 diagnosis of mild cognitive impairment in patients aged 50 and older. We developed prompts for two LLMs, Llama 2 and GPT-4, on Health Insurance Portability and Accountability Act (HIPAA)-compliant cloud-computing platforms using multiple approaches (e.g., hard prompting, retrieval augmented generation, and error analysis-based instructions) to select the optimal LLM-based method. Baseline models included a hierarchical attention-based neural network and XGBoost. Subsequently, we constructed an ensemble of the three models using a majority vote approach. Confusion-matrix-based scores were used for model evaluation. FindingsWe used a randomly annotated sample of 4,949 note sections from 1,969 patients (women: 1,046 [53.1%]; age: mean, 76.0 [SD, 13.3] years), filtered with keywords related to cognitive functions, for model development. For testing, a random annotated sample of 1,996 note sections from 1,161 patients (women: 619 [53.3%]; age: mean, 76.5 [SD, 10.2] years) without keyword filtering was utilised. GPT-4 demonstrated superior accuracy and efficiency compared to Llama 2, but did not outperform traditional models. The ensemble model outperformed the individual models in terms of all evaluation metrics with statistical significance (p<0.01), achieving a precision of 90.2% [95% CI: 81.9%-96.8%], a recall of 94.2% [95% CI: 87.9%-98.7%], and an F1-score of 92.1% [95% CI: 86.8%-96.4%]. Notably, the ensemble model showed a significant improvement in precision, increasing from a range of 70%-79% to above 90%, compared to the best-performing single model. Error analysis revealed that 63 samples were incorrectly predicted by at least one model; however, only 2 cases (3.2%) were mutual errors across all models, indicating diverse error profiles among them. InterpretationLLMs and traditional machine learning models trained using local EHR data exhibited diverse error profiles. The ensemble of these models was found to be complementary, enhancing diagnostic performance. Future research should investigate integrating LLMs with smaller, localised models and incorporating medical data and domain knowledge to enhance performance on specific tasks.

Low-light image enhancement: A comprehensive review on methods, datasets and evaluation metrics

Low-light image enhancement: A comprehensive review on methods, datasets and evaluation metrics

Remaining Service Life Prediction of Lithium-Ion Batteries Based on Randomly Perturbed Traceless Particle Filtering

To address the limitations in the prediction accuracy of the remaining useful life (RUL) of lithium-ion batteries, stemming from model accuracy, particle degradation, and insufficient diversity in the particle filter (PF) algorithm, this paper proposes a battery RUL prediction method utilizing a randomly perturbed unscented particle filter (RP-UPF) algorithm, based on the constructed battery capacity degradation model. The method utilizes evaluation metrics adjusted R-squared (Radj2) and the Akaike Information Criterion (AIC) to select the battery capacity decline model C5 with a higher goodness of fit. The initial values for constructing the C5 model are obtained using the relevance vector machine (RVM) and nonlinear least squares methods. Based on the constructed battery capacity decline model C5, the RP-UPF algorithm is employed to estimate the posterior parameters and iteratively approach the true battery capacity decline curve, thereby predicting the battery’s RUL. The research results indicate that, using battery B0005 as an example and starting the prediction from the 50th cycle, the RUL prediction results obtained with the RP-UPF algorithm demonstrate reductions in absolute error, relative error, and probability density function (PDF) width of 2%, 2.71%, and 10%, respectively, compared to the PF algorithm. Similar conclusions were drawn for batteries B0006 and B0018. Under the constructed battery capacity degradation model C5, the RP-UPF algorithm shows higher prediction accuracy for battery RUL and a narrower PDF range compared to the PF algorithm. This approach effectively addresses the issue of particle weight degradation in the PF algorithm, providing a more valuable reference for battery RUL prediction.

Hybrid real-time wave forecasting model combining Gaussian process regression and neural networks

To safely and effectively conduct offshore operations, a real-time wave forecast model that can accurately and rapidly forecast waves hours or days in advance is highly valuable. Recent studies have investigated models with neural networks (NNs) and have demonstrated their advantages in terms of forecasting speed and accuracy. However, NNs are limited by their requirement for a large training dataset, which can hinder the construction of the model if sufficient training data are unavailable. Therefore, this study constructed a hybrid real-time wave forecast model by combining an NN model and Gaussian process regression (GPR) model. The objective was to obtain reasonably accurate forecasts that were more accurate than those from individual NN models even when sufficient training data for the NN model was unavailable. To derive an optimal prediction, the values predicted by the NN and GPR models were combined by utilizing the standard deviation of the latter and considering the accuracy of each model. Consequently, predictions that were more accurate compared to those of the individual models for an entire year were obtained. This improvement was particularly significant for predicting relatively large wave heights, as indicated by the daily maximum wave heights. Additionally, various accuracy evaluation metrics demonstrated that despite using a small training dataset, the optimal prediction of the hybrid model was still superior to those of the individual models for predicting the daily maximum wave heights.

Earthquake magnitude prediction in Indonesia using a supervised method based on cloud radon data

In the challenging realm of earthquake prediction, the reliability of forecasting systems has remained a persistent obstacle. This study focuses on earthquake magnitude prediction in Indonesia, leveraging supervised machine learning techniques and cloud radon data. We present an analysis of the tele-monitoring system, data collection methods, and the application of regression-based machine learning algorithms. Utilizing a comprehensive dataset spanning 30 training instances and 105 test instances, the study evaluates multiple metrics to ascertain the efficacy of the prediction models. Our findings reveal that the linear regression approach yields the best earthquake magnitude prediction method, with the lowest values across multiple evaluation metrics: standard deviation 0.40, mean absolute error (MAE) 0.30, mean absolute percentage error (MAPE) 6%, root mean square error (RMSE) 0.52, mean squared error (MSE) 0.28, symmetric mean absolute percentage error (SMAPE) 0.06, and conformal normalized mean absolute percentage error (cnSMAPE) 0.97. Additionally, we discuss the implications of the research results and the potential applications in enhancing existing earthquake prediction methodologies.