Multiple Sampling Strategies Research Articles

Surrogate-Assisted Differential Evolution with multiple sampling mechanisms for high-dimensional expensive problems

Recently, Surrogate-Assisted Evolutionary Algorithms (SAEAs) have been widely employed in solving Expensive Optimization Problems (EOPs) due to their efficiency in obtaining satisfactory solutions with limited resources. By leveraging historical data to construct surrogate models for approximation, SAEAs can significantly reduce the number of expensive fitness evaluations for EOPs. A hierarchical SAEA optimization framework based on evolutionary sampling methods has achieved remarkable success in High-dimensional EOPs (HEOPs), which can effectively balance the exploration and exploitation capabilities. However, the majority of existing hierarchical SAEAs focus on either switching between different sampling strategies or enhancing a single sampling strategy, potentially overlooking the potential to improve multiple sampling strategies simultaneously. In this paper, we propose a Surrogate-Assisted Differential Evolution with Multiple Sampling Mechanisms (SADE-MSM) to tackle HEOPs, incorporating three sampling strategies with different mechanisms. The contributions of SADE-MSM are summarized as follows: 1) A centroid sampling method is applied before iterative optimization to enhance the early exploration ability; 2) An improved global prescreening sampling strategy is introduced to balance the exploration and exploitation capabilities; 3) A local search sampling with the adaptive optimal region strategy is proposed, significantly improving the exploitation ability. To validate the performance of SADE-MSM, we compared it with the state-of-the-art SAEAs on benchmark problems with dimensions ranging from 30 to 500. Experimental results demonstrate that SADE-MSM has a significant performance superiority.

Importations of SARS-CoV-2 lineages decline after nonpharmaceutical interventions in phylogeographic analyses

During the early stages of the SARS-CoV-2 pandemic, before vaccines were available, nonpharmaceutical interventions (NPIs) such as reducing contacts or antigenic testing were used to control viral spread. Quantifying their success is therefore key for future pandemic preparedness. Using 1.8 million SARS-CoV-2 genomes from systematic surveillance, we study viral lineage importations into Germany for the third pandemic wave from late 2020 to early 2021, using large-scale Bayesian phylogenetic and phylogeographic analysis with a longitudinal assessment of lineage importation dynamics over multiple sampling strategies. All major nationwide NPIs were followed by fewer importations, with the strongest decreases seen for free rapid tests, the strengthening of regulations on mask-wearing in public transport and stores, as well as on internal movements and gatherings. Most SARS-CoV-2 lineages first appeared in the three most populous states with most cases, and spread from there within the country. Importations rose before and peaked shortly after the Christmas holidays. The substantial effects of free rapid tests and obligatory medical/surgical mask-wearing suggests these as key for pandemic preparedness, given their relatively few negative socioeconomic effects. The approach relates environmental factors at the host population level to viral lineage dissemination, facilitating similar analyses of rapidly evolving pathogens in the future.

Opening a New Era with Machine Learning in Financial Services? Forecasting Corporate Credit Ratings Based on Annual Financial Statements

Corporate credit ratings provide multiple strategic, financial, and managerial benefits for decision-makers. Therefore, it is essential to have accurate and up-to-date ratings to continuously monitor companies’ financial situations when making financial credit decisions. Machine learning (ML)-based internal models can be used for the assessment of companies’ financial situations using annual statements. Particularly, it is necessary to check whether these ML models achieve better results compared to statistical methods. Due to the multi-class classification problem when forecasting corporate credit ratings, the development, monitoring, and maintenance of ML-based systems are more challenging compared to simple classifications. This problem becomes even more complex due to the required coordination with financial regulators (e.g., OECD, EBA, BaFin, etc.). Furthermore, the ML models must be updated regularly due to the periodic nature of annual statements as a dataset. To address the problem of the limited dataset, multiple sampling strategies and machine learning algorithms can be combined for accurate and up-to-date forecasting of credit ratings. This paper provides various implications for ML-based forecasting of credit ratings and presents an approach for combining sampling strategies and ML techniques. It also provides design recommendations for ML-based services in the finance industry on how to fulfill the existing regulations.

Reinforced PU-learning with Hybrid Negative Sampling Strategies for Recommendation

The data of recommendation systems typically only contain the purchased item as positive data and other un-purchased items as unlabeled data. To train a good recommendation model, in addition to the known positive information, we also need high-quality negative information. Capturing negative signals in positive and unlabeled data is challenging for recommendation systems. Most studies have used specific data and proposed negative sampling methods suitable to the data characteristics. Existing negative sampling strategies cannot automatically select suitable approaches for different data. However, this one-size-fits-all strategy often makes potential positive samples considered as negative, or truly negative samples considered as potential positive samples and recommend to users. In this way, it will not only turn down the recommendation result, but even also have an adverse effect. Accordingly, we propose a novel negative sampling model, Reinforced PU-learning with Hybrid Negative Sampling Strategies for Recommendation (RHNSR), which can combine multiple sampling strategies and dynamically adjust the proportions used by different sampling strategies. In addition, ensemble learning, which integrates various model sampling strategies for obtaining an improved solution, was applied to RHNSR. Extensive experiments were conducted on three real-world recommendation datasets, and the experimental results indicated that the proposed model significantly outperformed state-of-the-art baseline models and revealed significant improvements in precision and hit ratio (49.02% and 37.41%, respectively).

Quality Control Testing with Experimental Practical Illustrations under the Modified Lindley Distribution Using Single, Double, and Multiple Acceptance Sampling Plans

Quality control testing under acceptance sampling plans involves inspecting a representative sample of products or materials from a larger lot or batch to determine whether the lot meets predetermined quality standards. In this research, the modified Lindley distribution is used as a model for lifetime study. When a life test is amputated at a pre-appropriated time to decide on the admission or refusal of the submitted batches, the problems of the single, double, and multiple (three and four stages) acceptance sampling strategies are introduced. The optimal sample sizes are computed for single, double, and multiple acceptance sampling plans to ensure that the veritable mean life is greater than the prescribed mean life at the stipulated consumer’s risk. The operating characteristic functions are investigated at diverse quality levels. For single, double, and multiple acceptance sampling plans, the minimal ratios of the veritable mean life to the prescribed mean life at the established percent of the producer’s risk are obtained. To demonstrate the uses of single, double, and multiple, some numerical experiments are presented.

Application of data-driven surrogate models for active human model response prediction and restraint system optimization

Surrogate models are a must-have in a scenario-based safety simulation framework to design optimally integrated safety systems for new mobility solutions. The objective of this study is the development of surrogate models for active human model responses under consideration of multiple sampling strategies. A Gaussian process regression is chosen for predicting injury values based on the collision scenario, the occupant's seating position after a pre-crash movement and selected restraint system parameters. The trained models are validated and assessed for each sampling method and the best-performing surrogate model is selected for restraint system parameter optimization.

Diverse title generation for Stack Overflow posts with multiple-sampling-enhanced transformer

Stack Overflow is one of the most popular programming communities where developers can seek help for their encountered problems. Nevertheless, if inexperienced developers fail to describe their problems clearly, it is hard for them to attract sufficient attention and get the anticipated answers. To address such a problem, we propose M3NSCT5, a novel approach to automatically generate multiple post titles from the given code snippets. Developers may take advantage of the generated titles to find closely related posts and complete their problem descriptions. M3NSCT5 employs the CodeT5 backbone, which is a pre-trained Transformer model with an excellent language understanding and generation ability. To alleviate the ambiguity issue that the same code snippets could be aligned with different titles under varying contexts, we propose the maximal marginal multiple nucleus sampling strategy to generate multiple high-quality and diverse title candidates at a time for the developers to choose from. We build a large-scale dataset with 890,000 question posts covering eight programming languages to validate the effectiveness of M3NSCT5. The automatic evaluation results on the BLEU and ROUGE metrics demonstrate the superiority of M3NSCT5 over six state-of-the-art baseline models. Moreover, a human evaluation with trustworthy results also demonstrates the great potential of our approach for real-world applications.

HDP-TSRRT*: A Time–Space Cooperative Path Planning Algorithm for Multiple UAVs

This paper proposes a fast cooperative path planning algorithm for multiple UAVs that satisfies the time–space cooperative constraints, namely, the RRT* algorithm based on heuristic decentralized prioritized planning (HDP-TSRRT*), which takes into account the simultaneous arrival time variables of each UAV as well as the avoidance of conflicts and threats. HDP-TSRRT* is a hierarchical decoupling algorithm. First, all UAV pre-paths are planned simultaneously at the synchronous decentralized planning level. Second, at the coordination path level, the heuristic decentralized prioritized planning algorithm (HDP) is proposed to quickly complete the coordination process of the path planning sequence. This strategy assigns reasonable and robust priority to all UAVs based on the performance evaluation function composed of the number of potential collisions and the violation of collaboration time of the pre-planned path. Third, the time–space cooperative constraints-based RRT* algorithm (TSRRT*) is proposed at the single-machine cooperative path planning level. Based on this, the algorithm uses multiple sampling and cost evaluation strategies to guide the expansion of new nodes, and then optimizes neighborhood nodes based on the time coordination cost function so as to improve the efficiency of coordination path planning. Simulation and comparison show that HDP-TSRRT* has certain advantages in algorithm performance.

A selective evolutionary heterogeneous ensemble algorithm for classifying imbalanced data

<abstract> <p>Learning from imbalanced data is a challenging task, as with this type of data, most conventional supervised learning algorithms tend to favor the majority class, which has significantly more instances than the other classes. Ensemble learning is a robust solution for addressing the imbalanced classification problem. To construct a successful ensemble classifier, the diversity of base classifiers should receive specific attention. In this paper, we present a novel ensemble learning algorithm called Selective Evolutionary Heterogeneous Ensemble (SEHE), which produces diversity by two ways, as follows: 1) adopting multiple different sampling strategies to generate diverse training subsets and 2) training multiple heterogeneous base classifiers to construct an ensemble. In addition, considering that some low-quality base classifiers may pull down the performance of an ensemble and that it is difficult to estimate the potential of each base classifier directly, we profit from the idea of a selective ensemble to adaptively select base classifiers for constructing an ensemble. In particular, an evolutionary algorithm is adopted to conduct the procedure of adaptive selection in SEHE. The experimental results on 42 imbalanced data sets show that the SEHE is significantly superior to some state-of-the-art ensemble learning algorithms which are specifically designed for addressing the class imbalance problem, indicating its effectiveness and superiority.</p> </abstract>

A Fork in the Road for Emergency Medicine and Critical Care Blogs and Podcasts: Cross-sectional Study.

Free open-access meducation (FOAM) refers to open-access, web-based learning resources in medicine. It includes all formats of digital products, including blogs and podcasts. The number of FOAM blog and podcast sites in emergency medicine and critical care increased dramatically from 2002 to 2013, and physicians began to rely on the availability of these resources. The current landscape of these FOAM sites is unknown. This study aims to (1) estimate the current number of active, open-access blogs and podcasts in emergency medicine and critical care and (2) describe observed and anticipated trends in the FOAM movement using the Theory of Disruptive Innovation by Christensen as a theoretical framework. The authors used multiple resources and sampling strategies to identify active, open-access blogs and podcasts between April 25, 2022, and May 8, 2022, and classified these websites as blogs, podcasts, or blogs+podcasts. For each category, they reported the following outcome measures using descriptive statistics: age, funding, affiliations, and team composition. Based on these findings, the authors projected trends in the number of active sites using a positivist paradigm and the Theory of Disruptive Innovation as a theoretical framework. The authors identified 109 emergency medicine and critical care websites, which comprised 45.9% (n=50) blogs, 22.9% (n=25) podcasts, and 31.2% (n=34) blogs+podcasts. Ages ranged from 0 to 18 years; 27.5% (n=30) sold products, 18.3% (n=20) used advertisements, 44.0% (n=48) had institutional funding, and 27.5% (n=30) had no affiliation or external funding sources. Team sizes ranged from 1 (n=26, 23.9%) to ≥5 (n=60, 55%) individuals. There was a sharp decline in the number of emergency medicine and critical care blogs and podcasts in the last decade, dropping 40.4% since 2013. The initial growth of FOAM and its subsequent downturn align with principles in the Theory of Disruptive Innovation by Christensen. These findings have important implications for the field of medical education.

Distinct colon mucosa microbiomes associated with tubular adenomas and serrated polyps

Colorectal cancer is the second most deadly and third most common cancer in the world. Its development is heterogenous, with multiple mechanisms of carcinogenesis. Two distinct mechanisms include the adenoma-carcinoma sequence and the serrated pathway. The gut microbiome has been identified as a key player in the adenoma-carcinoma sequence, but its role in serrated carcinogenesis is less clear. In this study, we characterized the gut microbiome of 140 polyp-free and polyp-bearing individuals using colon mucosa and fecal samples to determine if microbiome composition was associated with each of the two key pathways. We discovered significant differences between the microbiomes of colon mucosa and fecal samples, with sample type explaining 10–15% of the variation observed in the microbiome. Multiple mucosal brushings were collected from each individual to investigate whether the gut microbiome differed between polyp and healthy intestinal tissue, but no differences were found. Mucosal aspirate sampling revealed that the microbiomes of individuals with tubular adenomas and serrated polyps were significantly different from each other and polyp-free individuals, explaining 1–4% of the variance in the microbiome. Microbiome composition also enabled the accurate prediction of subject polyp types using Random Forest, which produced an area under curve values of 0.87–0.99. By directly sampling the colon mucosa and distinguishing between the different developmental pathways of colorectal cancer, our study helps characterize potential mechanistic targets for serrated carcinogenesis. This research also provides insight into multiple microbiome sampling strategies by assessing each method’s practicality and effect on microbial community composition.

A Transformer-Based Feature Segmentation and Region Alignment Method for UAV-View Geo-Localization

Cross-view geo-localization is a task of matching the same geographic image from different views, e.g., unmanned aerial vehicle (UAV) and satellite. The most difficult challenges are the position shift and the uncertainty of distance and scale. Existing methods are mainly aimed at digging for more comprehensive fine-grained information. However, it underestimates the importance of extracting robust feature representation and the impact of feature alignment. The CNN-based methods have achieved great success in cross-view geo-localization. However it still has some limitations, e.g., it can only extract part of the information in the neighborhood and some scale reduction operations will make some fine-grained information lost. In particular, we introduce a simple and efficient transformer-based structure called Feature Segmentation and Region Alignment (FSRA) to enhance the model's ability to understand contextual information as well as to understand the distribution of instances. Without using additional supervisory information, FSRA divides regions based on the heat distribution of the transformer's feature map, and then aligns multiple specific regions in different views one on one. Finally, FSRA integrates each region into a set of feature representations. The difference is that FSRA does not divide regions manually, but automatically based on the heat distribution of the feature map. So that specific instances can still be divided and aligned when there are significant shifts and scale changes in the image. In addition, a multiple sampling strategy is proposed to overcome the disparity in the number of satellite images and that of images from other sources. Experiments show that the proposed method has superior performance and achieves the state-of-the-art in both tasks of drone view target localization and drone navigation. Code will be released at https://github.com/Dmmm1997/FSRA

BETA: a comprehensive benchmark for computational drug-target prediction.

Internal validation is the most popular evaluation strategy used for drug–target predictive models. The simple random shuffling in the cross-validation, however, is not always ideal to handle large, diverse and copious datasets as it could potentially introduce bias. Hence, these predictive models cannot be comprehensively evaluated to provide insight into their general performance on a variety of use-cases (e.g. permutations of different levels of connectiveness and categories in drug and target space, as well as validations based on different data sources). In this work, we introduce a benchmark, BETA, that aims to address this gap by (i) providing an extensive multipartite network consisting of 0.97 million biomedical concepts and 8.5 million associations, in addition to 62 million drug–drug and protein–protein similarities and (ii) presenting evaluation strategies that reflect seven cases (i.e. general, screening with different connectivity, target and drug screening based on categories, searching for specific drugs and targets and drug repurposing for specific diseases), a total of seven Tests (consisting of 344 Tasks in total) across multiple sampling and validation strategies. Six state-of-the-art methods covering two broad input data types (chemical structure- and gene sequence-based and network-based) were tested across all the developed Tasks. The best-worst performing cases have been analyzed to demonstrate the ability of the proposed benchmark to identify limitations of the tested methods for running over the benchmark tasks. The results highlight BETA as a benchmark in the selection of computational strategies for drug repurposing and target discovery.

Non-invasive cell-free DNA monitoring for predicting the response to neoadjuvant immunotherapy in locally advanced esophageal squamous cell carcinoma.

e16040 Background: The combined modality therapy including neoadjuvant chemoradiotherapy has been selected as the standard care for locally advanced Esophageal squamous cell carcinoma (ESCC). However, the conclusive evidences to support the effectiveness of neoadjuvant immunotherapy for ESCC are rarely discussed. This study aims to investigate the molecular features of cell-free DNA (cfDNA) and the impact of their dynamic changes on the response to neoadjuvant immunotherapy in locally advanced ESCC patients. Methods: Twenty-four eligible locally advanced ESCC patients with received neoadjuvant immunotherapy were enrolled in this study. 7 patients achieved pathological complete regression (pCR) and the other 17 patients had non-pCR. The baseline, second and third blood samples (before neoadjuvant immunotherapy, at one month and two months after neoadjuvant immunotherapy, respectively) were collected from Cancer Institute and Hospital, Chinese Academy of Medical Sciences. Both cfDNA fragments and genomic DNA were extracted for enrichment of a customized panel covering exon regions of 599 genes. The dynamic analysis of gene alterations, which allows the monitoring of the response to neoadjuvant immunotherapy in locally advanced ESCC patients, were conducted based on multiple blood sampling strategy. Results: The most frequently mutated genes among all cases were TP53, KMT2D, CREBBP and KMT2B. MSI scores of both second and third blood samples were remarkablely higher than those in baseline blood samples (p = 0.002, 0.04). On the contrary, the TMB values and mean variant allele frequency (VAF) of second and third blood samples were significantly decreased compared with the baseline blood (TMB, p = 0.04, 0.03; mean VAF, p = 0.01, 0.02). Moreover, a decreasing trend of mean VAF was observed at the time-series blood samples from ESCC patients with pCR, while the mean VAF of ESCC patients with non-pCR declined in the blood samples from the second time point then increased in the blood samples from the thrid time point. Conclusions: The detection of dynamic changes of mean VAF could distinguish the pCR and non-pCR ESCC patients with neoadjuvant immunotherapy. These findings warrant further expanded prospective cohorts to validate.

Optimizing stable isotope sampling design in terrestrial movement ecology research

AbstractThe recognition of adequate sampling designs is an interdisciplinary topic that has gained popularity over the last decades. In ecology, many research questions involve sampling across extensive and complex environmental gradients. This is the case for stable isotope analyses, which are widely used to characterize large‐scale movement patterns and dietary preferences of organisms across taxa. Because natural‐abundance stable isotope variation in the environment is incorporated into inert animal tissues, such as feathers or hair, it is possible to draw inferences about the type of food and water resources that individuals consumed and the locations where tissues were synthesized. However, modern stable isotope research can benefit from the implementation of robust statistical analyses and well‐designed sampling approaches to improve geographic assignment interpretation.We employed hydrogen stable isotope simulations to study inferences regarding the probability of origin of migratory individuals and reveal gaps in sampling efforts while highlighting uncertainties of assignment model extrapolations. We present an integrative approach that explores multiple sampling strategies across species with different geographic ranges to understand advantages and limitations of animal movement inferences based on stable isotope data.We show the characteristics of different sampling strategies through geographic and isotopic gradients and establish a set of diagnostic tools that uncover the attributes of these gradients and evaluate uncertainties of model results.Our analysis demonstrates that sampling regimes should be evaluated in relation to specific research questions and study constraints, and that adopting a single method across species ranges can lead to a costly but less effective sampling strategy.

Towards Robust Calculation of Interannual CO2 Growth Signal from TCCON (Total Carbon Column Observing Network)

The CO2 growth rate is one of the key geophysical quantities reflecting the dynamics of climate change as atmospheric CO2 growth is the primary driver of global warming. As recent studies have shown that TCCON (Total Carbon Column Observing Network) measurement footprints embrace quasi-global coverage, we examined the sensitivity of TCCON to the global CO2 growth. To this end, we used the aggregated TCCON observations (2006-2019) to retrieve Annual Growth Rate of CO2 (AGR) at global scales. The global AGR estimates from TCCON (AGRTCCON) are robust and independent, from (a) the station-wise seasonality, from (b) the differences in time series across the TCCON stations, and from (c) the type of TCCON stations used in the calculation (“background” or “contaminated” by neighboring CO2 sources). The AGRTCCON potential error, due to the irregular data sampling is relatively low (2.4–17.9%). In 2006–2019, global AGRTCCON ranged from the minimum of 1.59 ± 2.27 ppm (2009) to the maximum of 3.27 ± 0.82 ppm (2016), whereas the uncertainties express sub-annual variability and the data gap effects. The global AGRTCCON magnitude is similar to the reference AGR from satellite data (AGRSAT = 1.57–2.94 ppm) and the surface-based estimates of Global Carbon Budget (AGRGCB = 1.57–2.85). The highest global CO2 growth rate (2015/2016), caused by the record El Niño, was nearly perfectly reproduced by the TCCON (AGRTCCON = 3.27 ± 0.82 ppm vs. AGRSAT = 3.23 ± 0.50 ppm). The overall agreement between global AGRTCCON with the AGR references was yet weakened (r = 0.37 for TCCON vs. SAT; r = 0.50 for TCCON vs. GCB) due to two years (2008, 2015). We identified the drivers of this disagreement; in 2008, when only few stations were available worldwide, the AGRTCCON uncertainties were excessively high (AGRTCCON = 2.64 ppm with 3.92 ppm or 148% uncertainty). Moreover, in 2008 and 2015, the ENSO-driven bias between global AGRTCCON and the AGR references were detected. TCCON-to-reference agreement is dramatically increased if the years with ENSO-related biases (2008, 2015) are forfeited (r = 0.67 for TCCON vs. SAT, r = 0.82 for TCCON vs. GCB). To conclude, this is the first study that showed promising ability of aggregated TCCON signal to capture global CO2 growth. As the TCCON coverage is expanding, and new versions of TCCON data are being published, multiple data sampling strategies, dynamically changing TCCON global measurement footprint, and the irregular sensitivity of AGRTCCON to strong ENSO events; all should be analyzed to transform the current efforts into a first operational algorithm for retrieving global CO2 growth from TCCON data.

A Systematic Review of Multiple Linear Regression-Based Limited Sampling Strategies for Mycophenolic Acid Area Under the Concentration–Time Curve Estimation

Background and ObjectiveOne approach of therapeutic drug monitoring in the case of mycophenolic acid (MPA) is a limited sampling strategy (LSS), which allows the evaluation of the area under the concentration–time curve (AUC) based on few concentrations. The aim of this systematic review was to review the MPA LSSs and define the most frequent time points for MPA determination in patients with different indications for mycophenolate mofetil (MMF) administration.MethodsThe literature was comprehensively searched in July 2021 using PubMed, Scopus, and Medline databases. Original articles determining multiple linear regression (MLR)-based LSSs for MPA and its free form (fMPA) were included. Studies on enteric-coated mycophenolic sodium, previously established LSS, Bayesian estimator, and different than twice a day dosing were excluded. Data were analyzed separately for (1) adult renal transplant recipients, (2) adults with other than renal transplantation indication, and (3) for pediatric patients.ResultsA total of 27, 17, and 11 studies were found for groups 1, 2, and 3, respectively, and 126 MLR-based LSS formulae (n = 120 for MPA, n = 6 for fMPA) were included in the review. Three time-point equations were the most frequent. Four MPA LSSs: 2.8401 + 5.7435 × C0 + 0.2655 × C0.5 + 1.1546 × C1 + 2.8971 × C4 for adult renal transplant recipients, 1.783 + 1.248 × C1 + 0.888 × C2 + 8.027 × C4 for adults after islet transplantation, 0.10 + 11.15 × C0 + 0.42 × C1 + 2.80 × C2 for adults after heart transplantation, and 8.217 + 3.163 × C0 + 0.994 × C1 + 1.334 × C2 + 4.183 × C4 for pediatric renal transplant recipients, plus one fMPA LSS, 34.2 + 1.12 × C1 + 1.29 × C2 + 2.28 × C4 + 3.95 × C6 for adult liver transplant recipients, seemed to be the most promising and should be validated in independent patient groups before introduction into clinical practice. The LSSs for pediatric patients were few and not fully characterized. There were only a few fMPA LSSs although fMPA is a pharmacologically active form of the drug.ConclusionsThe review includes updated MPA LSSs, e.g., for different MPA formulations (suspension, dispersible tablets), generic form, and intravenous administration for adult and pediatric patients, and emphasizes the need of individual therapeutic approaches according to MMF indication. Five MLR-based MPA LSSs might be implemented into clinical practice after evaluation in independent groups of patients. Further studies are required, e.g., to establish fMPA LSS in pediatric patients.

The effect of measurement errors on the performance of the homogenously weighted moving average X¯ monitoring scheme with estimated parameters

ABSTRACT Classical monitoring schemes are typically designed under the assumption of known process parameters, perfect measurements and normality. In real-life applications, these assumptions are often violated. Thus, their Phase II performances are negatively affected by both measurement errors and parameter estimation. In this paper, the performance of the homogenously weighted moving average (HWMA) scheme is investigated under the assumption of unknown process parameters with and without measurement errors using the characteristics of the run-length distribution through intensive simulations. The negative effect of measurement errors is reduced using multiple measurements sampling strategy. It is found that the negative effect of the measurement errors is higher as the smoothing parameter increases and the larger the Phase I sample size, the smaller the negative effect of measurement errors. An illustrative example is given to demonstrate the implementation in real-life applications.

Reducing annotation effort in digital pathology: A Co-Representation learning framework for classification tasks.

Classification of digital pathology images is imperative in cancer diagnosis and prognosis. Recent advancements in deep learning and computer vision have greatly benefited the pathology workflow by developing automated solutions for classification tasks. However, the cost and time for acquiring high quality task-specific large annotated training data are subject to intra- and inter-observer variability, thus challenging the adoption of such tools. To address these challenges, we propose a classification framework via co-representation learning to maximize the learning capability of deep neural networks while using a reduced amount of training data. The framework captures the class-label information and the local spatial distribution information by jointly optimizing a categorical cross-entropy objective and a deep metric learning objective respectively. A deep metric learning objective is incorporated to enhance the classification, especially in the low training data regime. Further, a neighborhood-aware multiple similarity sampling strategy, and a soft-multi-pair objective that optimizes interactions between multiple informative sample pairs, is proposed to accelerate deep metric learning. We evaluate the proposed framework on five benchmark datasets from three digital pathology tasks, i.e., nuclei classification, mitosis detection, and tissue type classification. For all the datasets, our framework achieves state-of-the-art performance when using approximately only 50% of the training data. On using complete training data, the proposed framework outperforms the state-of-the-art on all the five datasets.

Molecular dynamics study on magnesium hydride nanoclusters with machine-learning interatomic potential

We introduce a machine-learning (ML) interatomic potential for Mg-H system based on Behler-Parrinello approach. In order to fit the complex bonding conditions in the cluster structure, we combine multiple sampling strategies to obtain training samples that contain a variety of local atomic environments. First-principles calculations based on density functional theory (DFT) are employed to get reference energies and forces for training the ML potential. For the calculation of bulk properties, phonon dispersion, gas-phase ${\mathrm{H}}_{2}$ interactions, and the potential energy surface (PES) for ${\mathrm{H}}_{2}$ dissociative adsorption on Mg(0001) surfaces, our ML potential has reached DFT accuracy at the level of GGA-PBE, and can be extended by combining the DFT-D3 method to describe van der Waals interaction. Moreover, through molecular dynamics (MD) simulations based on the ML potential, we find that for ${\mathrm{Mg}}_{n}{\mathrm{H}}_{m}$ clusters, $\mathrm{Mg}/\mathrm{Mg}{\mathrm{H}}_{x}$ phase separation occurs when $ml2n$, and for a cluster with a diameter of about 4 nm, the Mg part of the cluster forms a hexagonal close-packed (hcp) nanocrystalline structure at low temperature. Also, the calculated diffusion coefficients reproduce the experimental values and confirm an Arrhenius type temperature dependence in the range of 400 to 700 K.