Search

Targeted free energy perturbation uses an invertible mapping to promote configuration space overlap and the convergence of free energy estimates. However, developing suitable mappings can be challenging. Wirnsberger et al. [J. Chem. Phys. 153, 144112 (2020)] demonstrated the use of machine learning to train deep neural networks that map between Boltzmann distributions for different thermodynamic states. Here, we adapt their approach to the free energy differences of a flexible bonded molecule, deca-alanine, with harmonic biases and different spring centers. When the neural network is trained until "early stopping"—when the loss value of the test set increases—we calculate accurate free energy differences between thermodynamic states with spring centers separated by 1 Å and sometimes 2 Å. For more distant thermodynamic states, the mapping does not produce structures representative of the target state, and the method does not reproduce reference calculations. [ABSTRACT FROM AUTHOR]

ABSTRACT For phase II clinical trials that determine the acceptability of an experimental treatment based on ordinal toxicity and ordinal response, most monitoring methods require each ordinal outcome to be dichotomized using a selected cut‐point. This allows two early stopping rules to be constructed that compare marginal probabilities of toxicity and response to respective upper and lower limits. Important problems with this approach are loss of information due to dichotomization, dependence of treatment acceptability decisions on precisely how each ordinal variable is dichotomized, and ignoring association between the two outcomes. To address these problems, we propose a new Bayesian method, which we call U‐Bayes, that exploits elicited numerical utilities of the joint ordinal outcomes to construct one early stopping rule that compares the mean utility to a lower limit. U‐Bayes avoids the problems noted above by using the entire joint distribution of the ordinal outcomes, and not dichotomizing the outcomes. A step‐by‐step algorithm is provided for constructing a U‐Bayes rule based on elicited utilities and elicited limits on marginal outcome probabilities. A simulation study shows that U‐Bayes greatly improves the probability of determining treatment acceptability compared to conventional designs that use two monitoring rules based on marginal probabilities. [ABSTRACT FROM AUTHOR]

Ice avalanche (IA) has a strong concealment and sudden characteristics, which can cause severe disasters. The early identification of IA hidden danger is of great value for disaster prevention and mitigation. However, it is very difficult, and there is poor efficiency in identifying it by site investigation or manual remote sensing. So, an artificial intelligence method for the identification of IA hidden dangers using a deep learning model has been proposed, with the glacier area of the Yarlung Tsangpo River Gorge in Nyingchi selected for identification and validation. First, through engineering geological investigations, three key identification indices for IA hidden dangers are established, glacier source, slope angle, and cracks. Sentinel-2A satellite data, Google Earth, and ArcGIS are used to extract these indices and construct a feature dataset for the study and validation area. Next, key performance metrics, such as training accuracy, validation accuracy, test accuracy, and loss rates, are compared to assess the performance of the ResNet50 (Residual Neural Network 50) and VGG16 (Visual Geometry Group 16) models. The VGG16 model (96.09% training accuracy) is selected and optimized, using Early Stopping (ES) to prevent overfitting and L2 regularization techniques (L2) to add weight penalties, which constrained model complexity and enhanced simplicity and generalization, ultimately developing the ES-L2-VGG16 (Early Stopping—L2 Norm Regularization Techniques—Visual Geometry Group 16) model (98.61% training accuracy). Lastly, during the validation phase, the model is applied to the Yarlung Tsangpo River Gorge glacier area on the Tibetan Plateau (TP), identifying a total of 100 IA hidden danger areas, with average slopes ranging between 34° and 48°. The ES-L2-VGG16 model achieves an accuracy of 96% in identifying these hidden danger areas, ensuring the precise identification of IA dangers. This study offers a new intelligent technical method for identifying IA hidden danger, with clear advantages and promising application prospects. [ABSTRACT FROM AUTHOR]

Background: Orlistat is recommended as an adjunct to diet and exercise for weight loss in the treatment of type 2 diabetes mellitus (T2DM). Aim: To explore associations between patient characteristics and orlistat prescribing, and to determine associations of orlistat with weight loss in T2DM and prediabetes. Design and setting: Cohort study using anonymised health records from a UK database of general practice. Method: The UK Clinical Practice Research Datalink (CPRD) Aurum database was searched to compile a cohort of patients aged ≥18 years, first diagnosed with T2DM or prediabetes in 2016 or 2017. Once the data had been collated, multivariable logistic regression models were used to determine associations with starting orlistat and stopping it early (<12 weeks of prescriptions) and orlistat's associations with weight loss in those who had not been prescribed second-line antidiabetic medications. Results: Out of 100 552 patients with incident T2DM or prediabetes, 655 (0.8%) patients with T2DM and 128 (0.7%) patients with prediabetes were prescribed orlistat. Younger people, females, those in areas of deprivation, current smokers, those coprescribed metformin, and those recorded as having hypertension were statistically significantly more likely to be prescribed orlistat; higher baseline glycated haemoglobin levels were associated with early stopping. In comparison with patients not on orlistat, those who continued using it for ≥12 weeks were more likely to lose ≥5% weight (adjusted odds ratio [AOR] 1.69, 95% confidence interval [CI] = 1.07 to 2.67) but those who stopped orlistat early were less likely to lose ≥5% weight (AOR 0.56, 95% CI = 0.29 to 1.09). Conclusion: Orlistat was significantly associated with weight loss in patients with T2DM and prediabetes when taken for at least 12 weeks; however, it was infrequently prescribed and often taken for <12 weeks. Orlistat may be a useful adjunct to lifestyle modifications for patients with T2DM and prediabetes, but barriers to continued use means it may not be effective for everyone in managing weight loss. [ABSTRACT FROM AUTHOR]

AbstractNull hypothesis significance testing (NHST) is widely used in the field of human-computer interaction (HCI), and caution has been advised against deciding whether to stop or continue an experiment based on checking p-values. However, in clinical trials, NHST with interim analysis is commonly employed by adjusting the α levels to control Type I error rates. This approach, known as a group sequential design, can also be beneficial in the HCI field. We analyzed existing datasets of target-pointing tasks using group sequential designs and experimentally demonstrated that, depending on the α-level correction method, we could reduce the number of participants by 40.4–56.5%. Therefore, employing group sequential designs enables significant time and cost savings for both participants and researchers in the HCI field. [ABSTRACT FROM AUTHOR]

Background: For decades standard teaching recommended salt intake (sodium) reduction in patients suffering from heart failure. Neurohumoral activation with subsequent fluid retention provided a solid rationale for this long‐standing recommendation. Until recently no large randomized clinical trial of sodium restriction was available, while some observational studies and metanalyses even suggested a worse outcome with strict sodium restriction in patients with heart failure. Methods: In this narrative review we aimed to extricate from the literature whether strict sodium restriction is beneficial in patients with heart failure. We searched PubMed indexed articles between 2000 and 2023 for these terms: heart failure, salt, sodium, fluid intake. Results: Most randomized trials were small and showed a wide heterogeneity of interventions. A single large, randomized clinical trial was stopped early due to futility. Overall, there is no evidence that severe sodium restriction reduces the incidence of mortality and hospitalization in patients with heart failure. Quality of life and functional class may improve slightly with sodium restriction. Conclusion: Morbidity and mortality are not reduced with sodium restriction in patients with heart failure, although some symptomatic improvement may be expected. [ABSTRACT FROM AUTHOR]

Background Current guidance suggests oral antibiotics can be considered for children with acute otitis media (AOM) and ear discharge, but there is an absence of evidence regarding the relative effectiveness of antibiotic-corticosteroid eardrops. Aim To establish whether antibiotic-corticosteroid eardrops are non-inferior to oral antibiotics in children with AOM and ear discharge. Design and setting Open randomized controlled non-inferiority trial set in Dutch primary care. Methods Children were randomized to hydrocortisone-bacitracin-colistin eardrops (five drops, three times per day in the discharging ear(s)) or amoxicillin suspension (50 mg per kilogram of body weight per day, divided over three doses administered orally) for 7 days. The primary outcome was the proportion of children with resolution of ear pain and fever at day 3. Results Between December 2017 and March 2023, 58 of the planned 350 children were recruited due to slow accrual for various reasons. Children assigned to eardrops (n  = 26) had lower resolution rates of ear pain and fever at 3 days compared to those receiving oral antibiotics (n  = 31): 42% vs 65%; adjusted risk difference 20.3%, 95% confidence interval −5.3% to 41.9%), longer parent-reported ear discharge (6 vs 3 days; P  = .04), and slightly higher mean ear pain scores (Likert scale 0–6) over days 1–3 (2.1 vs 1.4, P  = .02), but received fewer oral antibiotic courses in 3months (11 for 25 children vs 33 for 30 children), and had less GI upset and rash (12% vs 32% and 8% vs 16%, respectively). Conclusion Early termination stopped us from determining non-inferiority of antibiotic-corticosteroid eardrops. Our limited data, requiring confirmation, suggest that oral antibiotics may be more effective than antibiotic-corticosteroid eardrops in resolving symptoms and shortening the duration of ear discharge. [ABSTRACT FROM AUTHOR]

A crucial factor in determining coal's quality is its gross calorific value (GCV). GCV is generally determined experimentally using an adiabatic bomb calorimeter. But the laboratory measurement process is complicated, tedious, expensive, and time-consuming. To address these problems, this study has developed five machine learning techniques, such as multilayer perceptron artificial neural network (MLP-ANN), multiple variable regression (MVR), adaptive neuro-fuzzy inference system (ANFIS), and support vector machine (SVM). ANN is further regularized using early stopping to avoid overfitting, whereas the MVR is regularized by employing L1, L2, and elastic networks. An inter-item correlation matrix, scatterplots, and sensitivity analysis are used to rank the parameters of the ultimate and proximate analyses according to their significance. A total of 6582 datapoints from the US. Geological Survey Coal Quality (COALQUAL) database are utilized for this study. The models' performance is measured using coefficient of determinant (${{\rm{R}}^2}$ R 2 ), mean squared error (MSE), explained error (EV), maximum error (max_error), minimum error (min_error), mean absolute error (MAE), and mean absolute percentage error (MAPE), respectively. The study's findings demonstrate that MVR, MLP-ANN, SVM, and ANFIS are highly efficient concerning both prediction and computation effectiveness. The prediction performances of these models are somewhat similar, with a coefficient of determinant of over 0.99. Hence, simple MVR is sufficient to predict GCV accurately instead of complex modeling. Parameters ranking indicates that oxygen (O), carbon (C), and moisture (M) are the most significant input variables for GCV prediction. [ABSTRACT FROM AUTHOR]

Value-based trials aim to maximize the expected net benefit by balancing technology adoption decisions and clinical trial costs. Adaptive trials offer additional efficiency. This article provides guidance on determining whether a value-based sequential design is the best option for an adaptive 2-arm trial, illustrated through a case study. We outlined 4 steps for the value-based sequential approach. The case study re-evaluates the Big CACTUS trial design using pilot trial data and a model-based health economic analysis. Expected net benefit is computed for (1) original fixed design, (2) value-based design with fixed sample size, and (3) optimal value-based sequential design with adaptive stopping. We compare pretrial modeling with the actual Big CACTUS trial results. Over 10 years, the adoption decision would affect approximately 215 378 patients. Pretrial modeling shows that the expected net benefit minus costs are (1) £102 million for the original fixed design, (2) £107 million (+5.3% higher) for the value-based design with optimal fixed sample size, and (3) £109 million (+6.7% higher) for the optimal value-based sequential design with maximum sample size of 435 per arm. Post hoc analysis using actual Big CACTUS trial data indicates that the value-adaptive trial with a maximum sample size of 95 participant pairs would not have stopped early. Bootstrap simulations reveal a 9.76% probability of early completion with n = 95 pairs compared with 31.50% with n = 435 pairs. The 4-step approach to value-based sequential 2-arm design with adaptive stopping was successfully implemented. Further application of value-based adaptive approaches could be useful to assess the efficiency of alternative study designs. • Value-based trials seek to balance the cost of a clinical trial with the expected health and economic benefits of the subsequent technology adoption decisions. Applying this approach to the adaptive design context has the potential to better balance the costs of clinical trials with the health benefits of health technology adoption decisions that are based on the results of those trials. • We bridge the gap between theory and practice by providing step-by-step guidance to implementing a value-based sequential design for an adaptive 2-arm trial. We demonstrate the first application of the value-based sequential design that uses data from an external pilot trial and do so in the context of publicly funded research. • With this tutorial, we hope to increase experience and application of these approaches so that they can be implemented with only small changes to current skills and practices. We also demonstrate mechanisms to assess the health economic value of doing a value-adaptive sequential trial compared with some other clinical trial designs. [ABSTRACT FROM AUTHOR]

As the scale of the power grid rapidly expands, its operation becomes increasingly complex, with higher demands on personnel proficiency, grid stability, equipment safety, and operational efficiency. In this study, a novel power insulator defect detection method based on convolutional neural networks (CNNs) is proposed. This method innovatively combines the feature extraction advantages of deep learning to build an efficient binary classification model capable of accurately detecting defects in power insulators in complex backgrounds. To avoid the impact of a small dataset on model performance, transfer learning was employed during model training to enhance the model's generalization ability. A combination of Grid Search and Random Search was used for hyperparameter tuning, and the Early Stopping strategy was introduced to effectively prevent the model from overfitting to the training set, ensuring generalization performance on the validation set. Experimental results show that the proposed method achieves an average accuracy of 98.6%, a recall of 96.8%, and an F1 score of 97.7% on the test set. Compared to traditional Faster RCNN and PCA-SVM methods, the proposed CNN model significantly improves detection accuracy and computational efficiency in complex backgrounds, exhibiting superior recognition precision and model generalization ability for efficiently and accurately identifying defective insulators. [ABSTRACT FROM AUTHOR]

This study focuses on the problem of corn yield prediction, and a novel prediction model based on a dynamic ensemble stacking regression algorithm is proposed. The model aims to achieve more accurate corn yield prediction based on the in-depth exploration of the potential correlations in multisource and multidimensional data. Data on the weather conditions, mechanization degree, and maize yield in Qiqihar City, Heilongjiang Province, from 1995 to 2022, are used. Important features are determined and extracted effectively by using principal component analysis and indicator contribution assessment methods. Based on the combination of an early stopping mechanism and parameter grid search optimization, the performance of eight base models, including a deep learning model, is fine-tuned. Based on the theory of heterogeneous ensemble learning, a threshold is established to stack the high-performing models, realizing a dynamic ensemble mechanism and employing averaging and optimized weighting methods for prediction. The results demonstrate that the prediction accuracy of the proposed dynamic ensemble regression model is significantly better as compared to the individual base models, with the mean squared error (MSE) being as low as 0.006, the root mean squared error (RMSE) being 0.077, the mean absolute error (MAE) being 0.061, and a high coefficient of determination value of 0.88. These findings not only validate the effectiveness of the proposed approach in the field of corn yield prediction but also highlight the positive role of multisource data fusion in enhancing the performance of prediction models. [ABSTRACT FROM AUTHOR]

Physiological signal recognition is crucial in emotion recognition, and recent advancements in multi-modal fusion have enabled the integration of various physiological signals for improved recognition tasks. However, current models for emotion recognition with hyper complexmulti-modal signals face limitations due to fusionmethods and insufficient attentionmechanisms, preventing further enhancement in classification performance. To address these challenges, we propose a new model framework named Signal Channel Attention Network (SCA-Net), which comprises three main components: an encoder, an attention fusion module, and a decoder. In the attention fusion module, we developed five types of attention mechanisms inspired by existing research and performed comparative experiments using the public dataset MAHNOB-HCI. All of these experiments demonstrate the effectiveness of the attention module we addressed for our baseline model in improving both accuracy and F1 score metrics. We also conducted ablation experiments within the most effective attention fusionmodule to verify the benefits of multi-modal fusion. Additionally, we adjusted the training process for different attention fusion modules by employing varying early stopping parameters to prevent model overfitting. [ABSTRACT FROM AUTHOR]

Objectives: In veterinary medicine, attempts to apply artificial intelligence (AI) to ultrasonography have rarely been reported, and few studies have investigated the value of AI in ultrasonographic diagnosis. This study aimed to develop a deep learning-based model for classifying the status of canine chronic kidney disease (CKD) using renal ultrasonographic images and assess its diagnostic performance in comparison with that of veterinary imaging specialists, thereby verifying its clinical utility. Materials and methods: In this study, 883 ultrasonograms were obtained from 198 dogs, including those diagnosed with CKD according to the International Renal Interest Society (IRIS) guidelines and healthy dogs. After preprocessing and labeling each image with its corresponding IRIS stage, the renal regions were extracted and classified based on the IRIS stage using the convolutional neural network-based object detection algorithm You Only Look Once. The training scenarios consisted of multi-class classification, categorization of images into IRIS stages, and four binary classifications based on specific IRIS stages. To prevent model overfitting, we balanced the dataset, implemented early stopping, used lightweight models, and applied dropout techniques. Model performance was assessed using accuracy, recall, precision, F1 score, and receiver operating characteristic curve and compared with the diagnostic accuracy of four specialists. Inter- and intra-observer variabilities among specialists were also evaluated. Results: The developed model exhibited a low accuracy of 0.46 in multi-class classification. However, a significant performance improvement was observed in binary classifications, with the model designed to distinguish stage 3 or higher showing the highest accuracy of 0.85. In this classification, recall, precision, and F1 score values were all 0.85, and the area under the curve was 0.89. Compared with radiologists, whose accuracy ranged from 0.48 to 0.62 in this experimental scenario, the AI model exhibited superiority. Intra-observer reliability among radiologists was substantial, whereas inter-observer variability showed a moderate level of agreement. Conclusions: This study developed a deep-learning framework capable of reliably classifying CKD IRIS stages 3 and 4 in dogs using ultrasonograms. The developed framework demonstrated higher accuracy than veterinary imaging specialists and provided more objective and consistent interpretations. Therefore, deep-learning-based ultrasound diagnostics are potentially valuable tools for diagnosing CKD in dogs. [ABSTRACT FROM AUTHOR]

Theoretical computational models are widely used to describe latent cognitive processes. However, these models do not equally explain data across participants, with some individuals showing a bigger predictive gap than others. In the current study, we examined the use of theory-independent models, specifically recurrent neural networks (RNNs), to classify the source of a predictive gap in the observed data of a single individual. This approach aims to identify whether the low predictability of behavioral data is mainly due to noisy decision-making or misspecification of the theoretical model. First, we used computer simulation in the context of reinforcement learning to demonstrate that RNNs can be used to identify model misspecification in simulated agents with varying degrees of behavioral noise. Specifically, both prediction performance and the number of RNN training epochs (i.e., the point of early stopping) can be used to estimate the amount of stochasticity in the data. Second, we applied our approach to an empirical dataset where the actions of low IQ participants, compared with high IQ participants, showed lower predictability by a well-known theoretical model (i.e., Daw's hybrid model for the two-step task). Both the predictive gap and the point of early stopping of the RNN suggested that model misspecification is similar across individuals. This led us to a provisional conclusion that low IQ subjects are mostly noisier compared to their high IQ peers, rather than being more misspecified by the theoretical model. We discuss the implications and limitations of this approach, considering the growing literature in both theoretical and data-driven computational modeling in decision-making science. [ABSTRACT FROM AUTHOR]

In several applications of text classification, training document labels are provided by human evaluators, and therefore, gathering sufficient data for model creation is time consuming and costly. The labeling time and effort may be reduced by active learning, in which classification models are created based on relatively small training sets, which are obtained by collecting class labels provided in response to labeling requests or queries. This is an iterative process with a sequence of models being fitted, and each of them is used to select query articles to be added to the training set for the next one. Such a learning scenario may pose different challenges for machine learning algorithms and text representation methods used for text classification than ordinary passive learning, since they have to deal with very small, often imbalanced data, and the computational expense of both model creation and prediction has to remain low. This work examines how classification algorithms and text representation methods that have been found particularly useful by prior work handle these challenges. The random forest and support vector machines algorithms are coupled with the bag of words and FastText word embedding representations and applied to datasets consisting of scientific article abstracts from systematic literature review studies in the biomedical domain. Several strategies are used to select articles for active learning queries, including uncertainty sampling, diversity sampling, and strategies favoring the minority class. Confidence-based and stability-based early stopping criteria are used to generate active learning termination signals. The results confirm that active learning is a useful approach to creating text classification models with limited access to labeled data, making it possible to save at least half of the human effort needed to assign relevant or irrelevant class labels to training articles. Two of the four examined combinations of classification algorithms and text representation methods were the most successful: the SVM algorithm with the FastText representation and the random forest algorithm with the bag of words representation. Uncertainty sampling turned out to be the most useful query selection strategy, and confidence-based stopping was found more universal and easier to configure than stability-based stopping. [ABSTRACT FROM AUTHOR]

Hierarchical Text Classification (HTC) aims to match text to hierarchical labels. Existing methods overlook two critical issues: first, some texts cannot be fully matched to leaf node labels and need to be classified to the correct parent node instead of treating leaf nodes as the final classification target. Second, error propagation occurs when a misclassification at a parent node propagates down the hierarchy, ultimately leading to inaccurate predictions at the leaf nodes. To address these limitations, we propose an uncertainty-guided HTC depth-aware model called DepthMatch. Specifically, we design an early stopping strategy with uncertainty to identify incomplete matching between text and labels, classifying them into the corresponding parent node labels. This approach allows us to dynamically determine the classification depth by leveraging evidence to quantify and accumulate uncertainty. Experimental results show that the proposed DepthMatch outperforms recent strong baselines on four commonly used public datasets: WOS (Web of Science), RCV1-V2 (Reuters Corpus Volume I), AAPD (Arxiv Academic Paper Dataset), and BGC. Notably, on the BGC dataset, it improves Micro-F1 and Macro-F1 scores by at least 1.09% and 1.74%, respectively. [ABSTRACT FROM AUTHOR]

Tooth decay is a dental condition characterized by the deterioration of tooth tissue originating from the outer surface and progressing to the pulp. Severe tooth decay, evolving into cavities, necessitates timely intervention to avert more serious dental-health issues. Common treatment procedures include filling and extraction of affected teeth. Presently, dentists conduct examinations for tooth decay by manually tallying affected, missing and filled teeth using an odontogram—a human tooth code diagram. This data is then recorded in patients' dental medical records. Recognizing the need for automation in assessing patients' experiences of tooth decay, this research endeavors to develop a model capable of detecting decayed, missing and filled teeth using variations of the YOLOv5 and YOLOv8 model architectures. The results of the training evaluation demonstrate the efficacy of YOLOv5l with a learning rate of 10-2, exhibiting a high precision value of 0.97, a recall of 0.858 and a mean average precision (mAP) of 0.904 within 1 hour and 18 minutes. According to the curves obtained in the training process, YOLOv5l shows great performance on the dental caries dataset, but precautions like early stopping are needed for a reliable and generalizable model. In contrast, YOLOv8 offers better training stability and larger variants perform better on the dental caries dataset, improving detection capabilities with continued training epochs. [ABSTRACT FROM AUTHOR]

AbstractPurposeMethodResultConclusionTo discuss whether early intervention can stop stuttering development. To inform junior clinicians and students of speech-language pathology about contemporary views on this issue.The issue was discussed by two university researchers and two speech-language pathologists who provide public clinical services. Written conversational turns in an exchange were limited to 100 words each. When that written dialogue was concluded, each participant provided 200 words of final reflection about the issue.Most differences that emerged centred on the clinical evidence base for early intervention, which emphasises stuttering reduction, and how it should be interpreted.The evidence base for early intervention has limitations and it should be interpreted cautiously. One interpretation is that reducing stuttering severity is a justifiable core of early management. Another interpretation focuses on ease of communication, anticipation of stuttering, and covert stuttering. [ABSTRACT FROM AUTHOR]

Summary Raw nanopore signals can be analyzed while they are being generated, a process known as real-time analysis. Real-time analysis of raw signals is essential to utilize the unique features that nanopore sequencing provides, enabling the early stopping of the sequencing of a read or the entire sequencing run based on the analysis. The state-of-the-art mechanism, RawHash, offers the first hash-based efficient and accurate similarity identification between raw signals and a reference genome by quickly matching their hash values. In this work, we introduce RawHash2, which provides major improvements over RawHash, including more sensitive quantization and chaining algorithms, weighted mapping decisions, frequency filters to reduce ambiguous seed hits, minimizers for hash-based sketching, and support for the R10.4 flow cell version and POD5 and SLOW5 file formats. Compared to RawHash, RawHash2 provides better F1 accuracy (on average by 10.57% and up to 20.25%) and better throughput (on average by 4.0× and up to 9.9×) than RawHash. Availability and implementation RawHash2 is available at https://github.com/CMU-SAFARI/RawHash. We also provide the scripts to fully reproduce our results on our GitHub page. [ABSTRACT FROM AUTHOR]

Discrete inverse problems correspond to solving a system of equations in a stable way with respect to noise in the data. A typical approach to select a meaningful solution is to introduce a regularizer. While for most applications the regularizer is convex, in many cases it is neither smooth nor strongly convex. In this paper, we propose and study two new iterative regularization methods, based on a primal-dual algorithm, to regularize inverse problems efficiently. Our analysis, in the noise free case, provides convergence rates for the Lagrangian and the feasibility gap. In the noisy case, it provides stability bounds and early stopping rules with theoretical guarantees. The main novelty of our work is the exploitation of some a priori knowledge about the solution set: we show that the linear equations determined by the data can be used more than once along the iterations. We discuss various approaches to reuse linear equations that are at the same time consistent with our assumptions and flexible in the implementation. Finally, we illustrate our theoretical findings with numerical simulations for robust sparse recovery and image reconstruction. We confirm the efficiency of the proposed regularization approaches, comparing the results with state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In rhinitis and asthma, several mHealth apps have been developed but only a few have been validated. However, these apps have a high potential for improving person‐centred care (PCC), especially in allergen immunotherapy (AIT). They can provide support in AIT initiation by selecting the appropriate patient and allergen shared decision‐making. They can also help in (i) the evaluation of (early) efficacy, (ii) early and late stopping rules and (iii) the evaluation of (carried‐over) efficacy after cessation of the treatment course. Future perspectives have been formulated in the first report of a joint task force (TF)—Allergic Rhinitis and Its Impact on Asthma (ARIA) and the European Academy of Allergy and Clinical Immunology (EAACI)—on digital biomarkers. The TF on AIT now aims to (i) outline the potential of the clinical applications of mHealth solutions, (ii) express their current limitations, (iii) make proposals regarding further developments for both clinical practice and scientific purpose and (iv) suggest which of the tools might best comply with the purpose of digitally‐enabled PCC in AIT. [ABSTRACT FROM AUTHOR]

Key Clinical Message: Despite the name, eosinophilia is not essential for diagnosing drug reaction with eosinophilia and systemic symptoms (DRESS syndrome). Early recognition and stopping the offending drug are vital to managing this condition, as it can otherwise lead to high mortality rates. [ABSTRACT FROM AUTHOR]

Fetal health care is vital in ensuring the health of pregnant women and the fetus. Regular check-ups need to be taken by the mother to determine the status of the fetus' growth and identify any potential problems. To know the status of the fetus, doctors monitor blood reports, Ultrasounds, cardiotocography (CTG) data, etc. Still, in this research, we have considered CTG data, which provides information on heart rate and uterine contractions during pregnancy. Several researchers have proposed various methods for classifying the status of fetus growth. Manual processing of CTG data is time-consuming and unreliable. So, automated tools should be used to classify fetal health. This study proposes a novel neural network-based architecture, the Dynamic Multi-Layer Perceptron model, evaluated from a single layer to several layers to classify fetal health. Various strategies were applied, including pre-processing data using techniques like Balancing, Scaling, Normalization hyperparameter tuning, batch normalization, early stopping, etc. to enhance the model's performance. A comparative analysis of the proposed method is done against the traditional machine learning models to showcase its accuracy (97%). An ablation study without any pre-processing techniques is also illustrated. This study easily provides valuable interpretations for healthcare professionals in the decision-making process. [ABSTRACT FROM AUTHOR]

Despite the widespread use of Cox regression for modeling treatment effects in clinical trials, in immunotherapy oncology trials and other settings therapeutic benefits are not immediately realized thereby violating the proportional hazards assumption. Weighted logrank tests and the so-called Maxcombo test involving the combination of multiple logrank test statistics have been advocated to increase power for detecting effects in these and other settings where hazards are nonproportional. We describe a testing framework based on supremum logrank statistics created by successively analyzing and excluding early events, or obtained using a moving time window. We then describe how such tests can be conducted in a group sequential trial with interim analyses conducted for potential early stopping of benefit. The crossing boundaries for the interim test statistics are determined using an easy-to-implement Monte Carlo algorithm. Numerical studies illustrate the good frequency properties of the proposed group sequential methods. [ABSTRACT FROM AUTHOR]

Summary: Arabidopsis Col‐0 RPP2A and RPP2B confer recognition of Arabidopsis downy mildew (Hyaloperonospora arabidopsidis [Hpa]) isolate Cala2, but the identity of the recognized ATR2Cala2 effector was unknown.To reveal ATR2Cala2, an F2 population was generated from a cross between Hpa‐Cala2 and Hpa‐Noks1. We identified ATR2Cala2 as a non‐canonical RxLR‐type effector that carries a signal peptide, a dEER motif, and WY domains but no RxLR motif. Recognition of ATR2Cala2 and its effector function were verified by biolistic bombardment, ectopic expression and Hpa infection.ATR2Cala2 is recognized in accession Col‐0 but not in Ler‐0 in which RPP2A and RPP2B are absent. In ATR2Emoy2 and ATR2Noks1 alleles, a frameshift results in an early stop codon. RPP2A and RPP2B are essential for the recognition of ATR2Cala2. Stable and transient expression of ATR2Cala2 under 35S promoter in Arabidopsis and Nicotiana benthamiana enhances disease susceptibility.Two additional Col‐0 TIR‐NLR (TNL) genes (RPP2C and RPP2D) adjacent to RPP2A and RPP2B are quantitatively required for full resistance to Hpa‐Cala2. We compared RPP2 haplotypes in multiple Arabidopsis accessions and showed that all four genes are present in all ATR2Cala2‐recognizing accessions. [ABSTRACT FROM AUTHOR]

Sequence alignment is an essential method in bioinformatics and the basis of many analyses, including phylogenetic inference, ancestral sequence reconstruction, and gene annotation. Sequencing artifacts and errors made during genome assembly, such as abiological frameshifts and incorrect early stop codons, can impact downstream analyses leading to erroneous conclusions in comparative and functional genomic studies. More significantly, while indels can occur both within and between codons in natural sequences, most amino-acid- and codon-based aligners assume that indels only occur between codons. This mismatch between biology and alignment algorithms produces suboptimal alignments and errors in downstream analyses. To address these issues, we present COATi, a statistical, codon-aware pairwise aligner that supports complex insertion–deletion models and can handle artifacts present in genomic data. COATi allows users to reduce the amount of discarded data while generating more accurate sequence alignments. COATi can infer indels both within and between codons, leading to improved sequence alignments. We applied COATi to a dataset containing orthologous protein-coding sequences from humans and gorillas and conclude that 41% of indels occurred between codons, agreeing with previous work in other species. We also applied COATi to semiempirical benchmark alignments and find that it outperforms several popular alignment programs on several measures of alignment quality and accuracy. [ABSTRACT FROM AUTHOR]

Most supervised learning methods require observation data and ground truth pairs as data sets to train the network. However, it is difficult and time-consuming to obtain a large number of high quality data sets, because ground truth is not available in some practical settings, such as medical imaging, dynamic scenes. Deep image prior (DIP) only uses one degraded image for image recovery tasks, which gets rid of the limitation of constructing a large number of training sets but requires an early stop mechanism. In order to further improve the image restoration ability of the DIP model, we combine it with the transformed total variation (TTV) regularization, which is a generalization of the classical total variation (TV) regularization and can achieve better performance for image restoration problems. The proposed method not only uses the deep neural network to capture image prior, but also exploits inherent sparse prior in image gradient domain. In addition, we provide an adaptive weight selection strategy for TTV regularization. The ADMM scheme is employed to solve the proposed models. Numerical results of image restoration illustrate that the proposed methods perform better than the compared methods. • A Transformed Total Variation and Deep Image prior algorithm is proposed for solving the image restoration task. • An adaptive weight selection strategy for TTV regularization is provided. • Numerical results of image restoration illustrate the efficiency of the proposed method among some state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Detecting Alzheimer’s disease (AD) using magnetic resonance imaging (MRI) is essential for early diagnosis and management. This study introduces a new method for detecting AD by combining three robust models: DenseNet201, EfficientNet B7, and extremely randomized trees (ERT). We improve the ability to extract features in DenseNet201 by including a self-attention mechanism. Additionally, we use early stopping techniques on EfficientNet B7 to address the issue of overfitting. In addition, Bayesian Optimization and Hyperband optimization techniques are used to adjust the hyperparameters of extra-trees to differentiate normal and abnormal MRI images. In addition, the authors used SHapley Additive exPlanations to understand the model’s decision. With minimal computer resources, the proposed model achieved a remarkable accuracy of 98.9% in detecting AD. The findings highlight the effectiveness of recommended feature extraction and ERT models and optimization methods to accurately identify AD using MRI images. [ABSTRACT FROM AUTHOR]

In coupled perturbed parameter ensemble (PPE) experiments or for development of a single coupled global climate model (GCM) in general, models can exhibit a slowdown in the Atlantic Meridional Overturning Circulation (AMOC) that can result in unrealistically reduced transport of heat and other tracers. Here we propose a method that researchers running PPE experiments can apply to their own PPE to diagnose what controls the AMOC strength in their model and make predictions thereof. As an example, using data from a 25-member coupled PPE experiment performed with HadGEM3-GC3.05, we found four predictors based on surface heat and freshwater fluxes in four critical regions from the initial decade of the spinup phase that could accurately predict the AMOC transport in the later stage of the experiment. The method, to our knowledge, is novel in that it separates the effects of the drivers of AMOC change from the effects of the changed AMOC. The identified drivers are shown to be physically credible in that the PPE members exhibiting AMOC weakening possess some combination of the following characteristics: warmer ocean in the North Atlantic Subpolar Gyre, fresher Arctic and Tropical North Atlantic Oceans and larger runoff from the Amazon and Orinoco Rivers. These characteristics were further traced to regional responses in atmosphere-only experiments. This study suggests promising potential for early stopping rules for parameter perturbations that could end up with an unrealistically weak AMOC, saving valuable computational resources. Some of the four drivers are likely to be relevant to other climate models so this study is of interest to model developers who do not have a PPE. [ABSTRACT FROM AUTHOR]

Underground mine structures like mine seals or built-in-place (BIP) refuge alternatives (RA) are exposed to specific loadings like explosions because of overlying and underlying strata conditions. Previous studies worked on the vertical loading response of structures like abutments from panel extraction but not the explosion resistance upon being subjected to these vertical strata loads. In this study, two steel-reinforced concrete wall designs as seal and RA applications are simulated to examine the performance and failure analysis under abutment-loading conditions during an explosion for a coal mine model using dynamic analysis by a distinct element code (3DEC). The available abutment monitoring data by longwall stoppings was used to estimate the abutment-loading conditions. The model setup was calibrated through two validation stages: (1) the concrete wall simulation with the explosion test of steel-reinforced concrete mine seal conducted by previous researchers; and (2) the stress change at the strata by the tailgate convergence-abutment-load multiplier curve. The calibrated models are subjected to various dynamic loadings to simulate explosions. The findings demonstrate that, in addition to magnitude, the pressure versus time mode of loading significantly influences the wall response. Instantaneous loading criteria can provide more cautious measurements to assess how well such walls operate in similar situations. Except for a few cases, the deformations after a single explosion are more pronounced for the effects of subsequent second and third explosions. As a result, the changes in deformation during subsequent explosions are relatively minor compared with the initial permanent deflections. [ABSTRACT FROM AUTHOR]

We propose a sequential, anytime-valid method to test the conditional independence of a response Y and a predictor X given a random vector Z. The proposed test is based on e-statistics and test martingales, which generalize likelihood ratios and allow valid inference at arbitrary stopping times. In accordance with the recently introduced model-X setting, our test depends on the availability of the conditional distribution of X given Z, or at least a sufficiently sharp approximation thereof. Within this setting, we derive a general method for constructing e-statistics for testing conditional independence, show that it leads to growth-rate optimal e-statistics for simple alternatives, and prove that our method yields tests with asymptotic power one in the special case of a logistic regression model. A simulation study is done to demonstrate that the approach is competitive in terms of power when compared to established sequential and nonsequential testing methods, and robust with respect to violations of the model-X assumption. for this article are available online. [ABSTRACT FROM AUTHOR]

Customer reviews inform potential buyers' decisions, but fake reviews in e-commerce can skew perceptions as customers may feel pressured to leave positive feedback. Detecting fake reviews in e-commerce platforms is a critical challenge, impacting online shopping and deceiving customers. Effective detection strategies, employing deep learning architectures and word embeddings, are essential to combat this issue. Specifically, the study presented in this paper employed a 1-layer Simple LSTM model, a 1D Convolutional model, and a combined CNN+LSTM model. These models were trained using different pre-trained word embeddings including Word2Vec, GloVe, FastText, and with Keras embeddings, to convert the text data into vector form. The models were evaluated based on accuracy and F1-score to provide a comprehensive measure of their performance. The results indicated that the Simple LSTM model with Word2Vec embeddings achieved an accuracy of nearly 91% and an F1-score of 0.9024, outperforming all other model-embedding combinations. The 1D convolutional model performed best without any embeddings, suggesting its ability to extract meaningful features from the raw text. The transformer-based models, BERT and DistilBERT, showed progressive learning but struggled with generalization, indicating the need for strategies such as early stopping, dropout, or regularization to prevent overfitting. Notably, the DistilBERT model consistently outperformed the LSTM model, achieving optimal performance with accuracy of 96% and an F1-score of 0.9639 using a batch size of 32 and a learning rate of 4.00E-05. [ABSTRACT FROM AUTHOR]

Common statistical theory applicable to confirmatory phase III trial designs usually assumes that patients are enrolled simultaneously and there is no time gap between enrollment and outcome observation. However, in practice, patients are enrolled successively and there is a lag between the enrollment of a patient and the measurement of the primary outcome. For single‐stage designs, the difference between theory and practice only impacts on the trial duration but not on the statistical analysis and its interpretation. For designs with interim analyses, however, the number of patients already enrolled into the trial and the number of patients with available outcome measurements differ, which can cause issues regarding the statistical analyses of the data. The main issue is that current methodologies either imply that at the time of the interim analysis there are so‐called pipeline patients whose data are not used to make a statistical decision (like stopping early for efficacy) or the enrollment into the trial needs to be at least paused for interim analysis to avoid pipeline patients. There are methods for delayed responses available that introduced error‐spending stopping boundaries for the enrollment of patients followed by critical values to reject the null hypothesis in case the stopping boundaries have been crossed beforehand. Here, we will discuss other solutions, considering different boundary determination algorithms using conditional power and introducing a design allowing for recruitment restart while keeping the type I error rate controlled. [ABSTRACT FROM AUTHOR]

This paper studies a distributionally robust multi-item newsvendor problem, where the demand distribution is unknown but specified with a general event-wise ambiguity set. Using the event-wise affine decision rules, we can obtain a conservative approximation formulation of the problem, which can typically be further reformulated as a linear program. In order to efficiently solve the resulting large-scale linear program, we develop a column generation-based decomposition scheme and speed up the computational efficiency by exploiting a special column selection strategy and stopping early based on a Karush-Kuhn-Tucker condition test. Focusing on the Wasserstein ambiguity set and the event-wise mean absolute deviation set, a computational study demonstrates both the computational efficiency of the proposed algorithm, which significantly outperforms a commercial solver and a Benders decomposition method, and the out-of-sample superiority of distributionally robust solutions relative to their sample average approximation counterparts. History: Accepted by Nicola Secomandi, Area Editor for Stochastic Models & Reinforcement Learning. Funding: This work was supported by the Natural Sciences and Engineering Research Council of Canada [492997-2016, RGPIN-2016-05208], the National Natural Science Foundation of China [71972012], Alliance de recherche numérique du Canada, and Canada Research Chairs [CRC-2018-00105]. It was also supported by Groupe d'études et de recherche en analyse des décisions (GERAD). Finally, this research was enabled in part by support provided by Digital Research Alliance of Canada (https://alliancecan.ca/en). Supplemental Material: The software that supports the findings of this study is available within the paper and its supplemental information (https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2022.0010) as well as from the IJOC GitHub software repository (https://github.com/INFORMSJoC/2022.0010). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/. [ABSTRACT FROM AUTHOR]

Bayesian predictive probabilities have become a ubiquitous tool for design and monitoring of clinical trials. The typical procedure is to average predictive probabilities over the prior or posterior distributions. In this paper, we highlight the limitations of relying solely on averaging, and propose the reporting of intervals or quantiles for the predictive probabilities. These intervals formalize the intuition that uncertainty decreases with more information. We present four different applications (Phase 1 dose escalation, early stopping for futility, sample size re-estimation, and assurance/probability of success) to demonstrate the practicality and generality of the proposed approach. [ABSTRACT FROM AUTHOR]

Introduction: Video laryngoscope (VL) technology improves first‐pass success. The novel i‐view VL device is inexpensive and disposable. We sought to determine the first‐pass intubation success with the i‐view VL device versus the standard reusable VL systems in routine use at each site. Methods: We performed a prospective, pragmatic study at two major emergency departments (EDs) when VL was used. We rotated i‐view versus reusable VL as the preferred device of the month based on an a priori schedule. An investigator‐initiated interim analysis was performed. Our primary outcome was a first‐pass success with a non‐inferiority margin of 10% based on the per‐protocol analysis. Results: There were 93 intubations using the reusable VL devices and 81 intubations using the i‐view. Our study was stopped early due to futility in reaching our predetermined non‐inferiority margin. Operator and patient characteristics were similar between the two groups. The first‐pass success rate for the i‐view group was 69.1% compared to 84.3% for the reusable VL group. A non‐inferiority analysis indicated that the difference (−15.1%) and corresponding 90% confidence limits (−25.3% to −5.0%) did not fall within the predetermined 10% non‐inferiority margin. Conclusions: The i‐view device failed to meet our predetermined non‐inferiority margin when compared to the reusable VL systems with the study stopping early due to futility. Significant crossover occurred at the discretion of the intubating operator during the i‐view month. [ABSTRACT FROM AUTHOR]

Ransomware poses a significant threat by encrypting files or systems demanding a ransom be paid. Early detection is essential to mitigate its impact. This paper presents an Uncertainty-Aware Dynamic Early Stopping (UA-DES) technique for optimizing Deep Belief Networks (DBNs) in ransomware detection. UA-DES leverages Bayesian methods, dropout techniques, and an active learning framework to dynamically adjust the number of epochs during the training of the detection model, preventing overfitting while enhancing model accuracy and reliability. Our solution takes a set of Application Programming Interfaces (APIs), representing ransomware behavior as input we call "UA-DES-DBN". The method incorporates uncertainty and calibration quality measures, optimizing the training process for better more accurate ransomware detection. Experiments demonstrate the effectiveness of UA-DES-DBN compared to more conventional models. The proposed model improved accuracy from 94% to 98% across various input sizes, surpassing other models. UA-DES-DBN also decreased the false positive rate from 0.18 to 0.10, making it more useful in real-world cybersecurity applications. [ABSTRACT FROM AUTHOR]

Physical experiments are often expensive and time-consuming. Test engineers must certify the compatibility of aircraft and their weapon systems before they can be deployed in the field, but the testing required is time consuming, expensive, and resource limited. Adopting Bayesian adaptive designs is a promising way to borrow from the successes seen in the clinical trials domain. The use of predictive probability (PP) to stop testing early and make faster decisions is particularly appealing given the aforementioned constraints. Given the high-consequence nature of the tests performed in the national security space, a strong understanding of new methods is required before being deployed. Although PP has been thoroughly studied for binary data, there is less work with continuous data, where many reliability studies are interested in certifying the specification limits of components. A simulation study evaluating the robustness of this approach indicates early stopping based on PP is reasonably robust to minor assumption violations, especially when only a few interim analyses are conducted. The simulation study also compares PP to conditional power, showing its relative strengths and weaknesses. A post-hoc analysis exploring whether release requirements of a weapon system from an aircraft are within specification with desired reliability resulted in stopping the experiment early and saving 33% of the experimental runs. [ABSTRACT FROM AUTHOR]

The urban rail transit network has gradually realized grid operation with the increase in the coverage rate. Therefore, the stopping schemes in accordance with the trend of the passenger flow are more conducive to improving the attractiveness of the rail transit and improving the sharing rate of the urban public transit. Traditional data from a single source may not be sufficient to describe the overall trend of the passenger flow in a period of time, and the error is possible in the case of insufficient data. Based on the multisource data, the spatial weight function is introduced to fuse the point of interest data and real estate information data, from which one obtains the residential index and office index, and the cluster analysis is conducted to obtain the potential stop scheme. Then, the optimization model of the train operation plan is established, aiming at minimizing the passenger travel time and the generalized system cost, and is constrained by a series of driving conditions. Compared with the single data source, multisource data can better reflect passenger flow trends and land use characteristics. Compared with the traditional all-station stopping scheme, a reasonable setting of crossing stations and running express-local trains can better satisfy the demands of the passenger flow. Finally, the optimization of Changchun rail Transit Line 1 shows that the model can reduce the travel time of passengers and the operating cost of the rail transit company and improve the quality of service, so as to achieve a win-win situation between passengers and the rail transit company. [ABSTRACT FROM AUTHOR]

Background: Classification of binary data arises naturally in many clinical applications, such as patient risk stratification through ICD codes. One of the key practical challenges in data classification using machine learning is to avoid overfitting. Overfitting in supervised learning primarily occurs when a model learns random variations from noisy labels in training data rather than the underlying patterns. While traditional methods such as regularization and early stopping have demonstrated effectiveness in interpolation tasks, addressing overfitting in the classification of binary data, in which predictions always amount to extrapolation, demands extrapolation-enhanced strategies. One such approach is hybrid mechanistic/data-driven modeling, which integrates prior knowledge on input features into the learning process, enhancing the model's ability to extrapolate. Results: We present NoiseCut, a Python package for noise-tolerant classification of binary data by employing a hybrid modeling approach that leverages solutions of defined max-cut problems. In a comparative analysis conducted on synthetically generated binary datasets, NoiseCut exhibits better overfitting prevention compared to the early stopping technique employed by different supervised machine learning algorithms. The noise tolerance of NoiseCut stems from a dropout strategy that leverages prior knowledge of input features and is further enhanced by the integration of max-cut problems into the learning process. Conclusions: NoiseCut is a Python package for the implementation of hybrid modeling for the classification of binary data. It facilitates the integration of mechanistic knowledge on the input features into learning from data in a structured manner and proves to be a valuable classification tool when the available training data is noisy and/or limited in size. This advantage is especially prominent in medical and biomedical applications where data scarcity and noise are common challenges. The codebase, illustrations, and documentation for NoiseCut are accessible for download at https://pypi.org/project/noisecut/. The implementation detailed in this paper corresponds to the version 0.2.1 release of the software. [ABSTRACT FROM AUTHOR]

Aims: Computerised adaptive test (CAT) provides individualised patient reported outcome measurement while retaining direct comparability of scores across patients and studies. Optimal CAT measurement requires an appropriate CAT-setting, the set of criteria defining the CAT including start item, item selection criterion, and stop criterion. The European Organisation for Research and Treatment of Cancer (EORTC) CAT Core allows for assessing the 14 functional and symptom domains covered by the EORTC QLQ-C30 questionnaire. The aim was to present a general approach for selecting CAT-settings and to use this to develop a portfolio of standard settings for the EORTC CAT Core optimised for different purposes and populations. Methods: Using simulations, the measurement properties of CATs of different length and precision were evaluated and compared allowing for identifying the most suitable settings. All CATs were initiated with the most informative QLQ-C30 item. For each domain two fixed-length and two fixed-precision standard CATs were selected focusing on efficiency (brief version) and precision (long), respectively. Results: The brief fixed-length CATs included 3–5 items each while the long versions included 5–8 items. The fixed-precision CATs aimed for reliability of 0.65–0.95 (brief versions) and 0.85–0.98 (long versions), respectively. Median sample size savings using the CATs compared to the QLQ-C30 scales ranged 20%-31%, although savings varied considerably across the domains. Conclusion: The EORTC CAT Core standard settings simplify selection of relevant and appropriate CATs. The CATs prioritise either brevity and efficiency or precision, but all provide increased measurement precision and hence, reduced sample size requirements compared to the QLQ-C30 scales. The CATs may be used as they are or modified to accommodate specific requirements. [ABSTRACT FROM AUTHOR]

The paper presents a review of the current state of subgradient and accelerated convex optimization methods, including the cases with the presence of noise and access to various information about the objective function (function value, gradient, stochastic gradient, higher derivatives). For nonconvex problems, the Polyak–Lojasiewicz condition is considered and a review of the main results is given. The behavior of numerical methods in the presence of a sharp minimum is considered. The aim of this review is to show the influence of the works of B.T. Polyak (1935–2023) on gradient optimization methods and their surroundings on the modern development of numerical optimization methods. [ABSTRACT FROM AUTHOR]

Due to highly unmet medical needs in rare disease areas, there is great desire to speed up the drug development process. A two-stage adaptive design option for a placebo-controlled registration study is being evaluated. Stage 1 consists of participants in an ongoing phase 2 study with 1-year double blinded (DB) treatment and Stage 2 includes newly enrolled participants with 2-year DB treatment period. The primary endpoint is the annualized rate of change (slope) for a continuous longitudinal measurement, which will be evaluated through a random coefficient linear mixed model. An unblinded interim analysis will be performed using Stage 1 data to re-estimate the sample size for Stage 2, followed by another interim analysis for potential early efficacy stopping when all participants completed the 1-year DB treatment. To control the overall Type 1 error rate, rather than using a conservative approach, the actual correlation between the interim and final test statistics will be taken into account to determine the final significance level given the pre-specified significance level for the interim efficacy analysis. Multiplicity adjustments for secondary endpoints including considerations for order switching between interim and final analysis in this specific case study will also be discussed. [ABSTRACT FROM AUTHOR]

Background: Group sequential designs incorporating the option to stop for futility at the time point of an interim analysis can save time and resources. Thereby, the choice of the futility boundary importantly impacts the design's resulting performance characteristics, including the power and probability to correctly or wrongly stop for futility. Several authors contributed to the topic of selecting good futility boundaries. For binary endpoints, Simon's designs (Control Clin Trials 10:1–10, 1989) are commonly used two-stage designs for single-arm phase II studies incorporating futility stopping. However, Simon's optimal design frequently yields an undesirably high probability of falsely declaring futility after the first stage, and in Simon's minimax design often a high proportion of the planned sample size is already evaluated at the interim analysis leaving only limited benefit in case of an early stop. Methods: This work focuses on the optimality criteria introduced by Schüler et al. (BMC Med Res Methodol 17:119, 2017) and extends their approach to binary endpoints in single-arm phase II studies. An algorithm for deriving optimized futility boundaries is introduced, and the performance of study designs implementing this concept of optimal futility boundaries is compared to the common Simon's minimax and optimal designs, as well as modified versions of these designs by Kim et al. (Oncotarget 10:4255–61, 2019). Results: The introduced optimized futility boundaries aim to maximize the probability of correctly stopping for futility in case of small or opposite effects while also setting constraints on the time point of the interim analysis, the power loss, and the probability of stopping the study wrongly, i.e. stopping the study even though the treatment effect shows promise. Overall, the operating characteristics, such as maximum sample size and expected sample size, are comparable to those of the classical and modified Simon's designs and sometimes better. Unlike Simon's designs, which have binding stopping rules, the optimized futility boundaries proposed here are not adjusted to exhaust the full targeted nominal significance level and are thus still valid for non-binding applications. Conclusions: The choice of the futility boundary and the time point of the interim analysis have a major impact on the properties of the study design. Therefore, they should be thoroughly investigated at the planning stage. The introduced method of selecting optimal futility boundaries provides a more flexible alternative to Simon's designs with non-binding stopping rules. The probability of wrongly stopping for futility is minimized and the optimized futility boundaries don't exhibit the unfavorable properties of an undesirably high probability of falsely declaring futility or a high proportion of the planned sample evaluated at the interim time point. [ABSTRACT FROM AUTHOR]

For the attitude control of hypersonic vehicles, a control scheme based on linear active disturbance rejection (LADRC) and modified deep deterministic policy gradient (MDDPG) is proposed. Firstly, LADRC is used to deal with uncertainty and nonlinear problems in the attitude control process. For the tedious manual parameter tuning process, MDDPG is used to optimize the control gains and bandwidth of LADRC. Secondly, a modified reward function and an early stop criterion are introduced in the MDDPG algorithm to improve the optimization performance. Then, another MDDPG is used as an auxiliary control, which is combined with LADRC for attitude control to improve the robustness and accuracy of the control. The proposed method considers the robustness and rapidity of the control. Finally, the effectiveness of the proposed method is proved by simulation. Compared with traditional LADRC, LADRC based on the Q-learning algorithm, LADRC based on the traditional DDPG algorithm, and LADRC based on MDDPG algorithm, the proposed method has a better control effect and can avoid a lot of manual parameter tuning. [ABSTRACT FROM AUTHOR]

Prior aquatic animal image classification research focused on distinguishing external features in controlled settings, utilizing either digital cameras or webcams. Identifying visually similar species, like Short mackerel (Rastrelliger brachysoma) and Indian mackerel (Rastrelliger kanagurta), is challenging without specialized knowledge. However, advancements in computer technology have paved the way for leveraging machine learning and deep learning systems to address such challenges. In this study, transfer learning techniques were employed, utilizing established pre-trained models such as ResNet50, Xception, InceptionV3, VGG19, VGG16, and MobileNetV3Small. These models were applied to differentiate between the two species using raw images captured by a smartphone under uncontrolled conditions. The core architecture of the pre-trained models remained unchanged, except for the removal of the final fully connected layer. Instead, a global average pooling layer and two dense layers were appended at the end, comprising 1024 units and by a single unit, respectively. To mitigate overfitting concerns, early stopping was implemented. The results revealed that, among the models assessed, the Xception model exhibited the most promising predictive performance. It achieved the highest average accuracy levels of 0.849 and 0.754 during training and validation, surpassing the other models. Furthermore, fine-tuning the Xception model by extending the number of epochs yielded more impressive outcomes. After 30 epochs of fine-tuning, the Xception model demonstrated optimal performance, reaching an accuracy of 0.843 and displaying a 11.508% improvement in predictions compared to the model without fine-tuning. These findings highlight the efficacy of transfer learning, particularly with the Xception model, in accurately distinguishing visually similar aquatic species using smartphone-captured images, even in uncontrolled conditions. [ABSTRACT FROM AUTHOR]

Targets with rotating components generate micro‐motion (MM) modulation effect in addition to the main body. Extracting MM parameters is challenging due to interference from the target's main body, necessitating the separation of modulation signals. This letter proposes a robust complex local mean decomposition (RCLMD) method with self‐adaptive sifting stopping, aiming at the problem of component redundancy due to multiple iterations during break and the loss of modulation components during the separation process. The proposed method sets the objective function and self‐adaptive stopping criterion, combined with the modulation signal characteristics, enhancing the accuracy and efficiency of MM component extraction. Simulation experiments show that compared with the complex local mean decomposition method, the complex empirical mode decomposition method, and its improved method, the RCLMD method can achieve the highest decomposition effect of 96.57%, and the separation time consumed has a significant advantage over the above methods, performance is less fluctuating by the change of signal‐to‐noise ratio with good robustness. The measured data in real scenarios also verify the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]