Number Of Samples Research Articles

What can be learnt with wide convolutional neural networks?*

Abstract Understanding how convolutional neural networks (CNNs) can efficiently learn high-dimensional functions remains a fundamental challenge. A popular belief is that these models harness the local and hierarchical structure of natural data such as images. Yet, we lack a quantitative understanding of how such structure affects performance, for example the rate of decay of the generalisation error with the number of training samples. In this paper, we study infinitely wide deep CNNs in the kernel regime. First, we show that the spectrum of the corresponding kernel inherits the hierarchical structure of the network, and we characterise its asymptotics. Then, we use this result together with generalisation bounds to prove that deep CNNs adapt to the spatial scale of the target function. In particular, we find that if the target function depends on low-dimensional subsets of adjacent input variables then the decay of the error is controlled by the effective dimensionality of these subsets. Conversely, if the target function depends on the full set of input variables then the error decay is controlled by the input dimension. We conclude by computing the generalisation error of a deep CNN trained on the output of another deep CNN with randomly initialised parameters. Interestingly, we find that, despite their hierarchical structure, the functions generated by infinitely wide deep CNNs are too rich to be efficiently learnable in high dimensions.

Semi-supervised comparative learning compensation method for chemical gas sensor drift.

The gradual and unpredictable variation in chemo-sensory signal responses when exposed to the same analyte under identical conditions, commonly referred to as sensor drift, has long been recognized as one of the most serious challenges faced by chemical sensors. The traditional drift compensation method is both labor-intensive and expensive, as it requires frequent collection and labeling of gas samples for recalibration. Introducing a small number of meaningful drift calibration samples can be an attractive strategy to reduce the computational load and improve the performance of the updated classifier. However, under the influence of drift, new challenges arise due to the difference in the distribution of source and target domain data. This paper proposes a novel algorithm framework called semi-supervised contrastive learning drift compensation (SSCLDC). The framework automatically extracts high-level abstract features based on a multilayer perceptron to better represent the structure of the source data. In addition, to address the issue of data distribution differences caused by drift between the source and target domains. We add a small number of reference sample pairs into the training for semi-supervised learning. Combining a contrastive loss function that can represent the matching degree of paired samples effectively overcomes the problem of sensor drift. The Kennard-Stone sequential algorithm is used to select the representative reference sample from the set of candidate reference samples. Experiments conducted on a widely used long-term chemical gas sensor drift dataset demonstrate that the proposed method outperforms several classic drift compensation techniques, highlighting its effectiveness and practical applicability.

Sparse cross-transformer network for surface defect detection.

Quality control processes with automation ensure that customers receive defect-free products that meet their needs. However, the performance of real-world surface defect detection is often severely hindered by the scarcity of data. Recently, few-shot learning has been widely proposed as a solution to the data sufficiency problem by leveraging a limited number of base class samples. However, achieving discriminative and generalization capabilities with few samples remains a challenging task in various surface defect detection scenarios. In this paper, we propose a sparse cross-transformer network (SCTN) for surface defect detection. Specifically, we introduce a residual layer module to enhance the network's ability to retain crucial information. Next, we propose a sparse layer module within the cross-transformer to increase computational efficiency. Finally, we incorporate a squeeze-and-excitation network into the cross-transformer to enhance the attention mechanism between local patches outputted by the transformer encoder. To verify the effectiveness of our proposed method, we conducted extensive experiments on the cylinder liner defect dataset, the NEU steel surface defect dataset, and the PKU-Market-PCB dataset, achieving the best mean average precision of 62.73%, 85.29%, and 88.7%, respectively. The experimental results demonstrate that our proposed method achieves significant improvements compared to state-of-the-art algorithms. Additionally, the results indicate that SCTN enhances the network's discriminative ability and effectively improves generalization across various surface defect detection tasks.

Numerical Study on Evaluation of Environmental DNA Approach for Estimating Fish Abundance and Distribution in Semi-Enclosed Bay

Despite efforts to use environmental DNA (eDNA), accurately quantifying fish populations remains a challenge. A recent eDNA approach provided reliable estimates of coastal fish population abundance, but it was not as effective for assessing spatial distribution due to a lack of eDNA samples relative to the study area. Therefore, we conducted a numerical case study to evaluate the ability of the eDNA approach to estimate fish (Jack mackerel) abundance and distribution based on the number of eDNA samples in a semi-enclosed bay (Jinhae Bay). Our study revealed that the eDNA approach can provide reliable estimates of fish abundance, even with knowledge of the eDNA concentration in just 1% of the study area. However, for estimating spatial distribution and fish school, significant estimates were obtained only when the eDNA concentration was identified in more than 70% of the study area. Our results confirm that the eDNA approach can reflect fish abundance but has limitations in estimating fish distribution.

Extended-spectrum beta-laktamase (ESBL) enzyme producing bacteria isolated from rectal swabs of around animals in dairy farms

This study aims to identify bacterial contamination of Klebsiella pneumoniae and Escherichia coli producing Extended Spectrum Beta Lactamase (ESBL) isolated from rectal swabs of local animals on a dairy farm in Deyeng Village, Kediri, East Java. Interviews and observations were conducted with farmers to obtain the required risk factor data (cleanliness of drinking and eating places, drinking water sources, and administration of antibiotics to livestock). The number of samples tested in this study was 34 samples, including 12 chicken samples, 8 goat samples, 5 duck samples, 4 rabbit samples, 3 cat samples and 2 goose samples. Samples were isolated on MCA media + cefotaxime 1 mg/l and isolated again on MCA media without the addition of antibiotics to multiply pure bacterial colonies. The tests carried out to determine the presence of ESBL-producing bacteria are biochemical tests, namely the IMViC test (Indole, Methyl red, Voges-Proskauer, Citrate Utilization), TSIA (Triple Sugar Iron Agar) test, gram staining, and the Double Disk Standard Method test using β-lactam disc antibiotics, namely cefotaxime 30μg, ceftazidime 30μg, and amoxycilline-clavulanate acid 30μg. The results of this research were that 6 samples or 17.6% of ESBL-producing bacteria were identified from poultry groups (chickens, ducks and geese) from dairy farms in Deyeng Village, Kediri Regency, East Java.

Lack of Association Between BsmI and FokI Polymorphisms of the VDR Gene and Sporadic Colorectal Cancer in a Romanian Cohort-A Preliminary Study.

Colorectal cancer (CRC) is a major public health problem worldwide, currently ranking third in cancer incidence and second in mortality. Multiple genes and environmental factors have been involved in the complex and multifactorial process of CRC carcinogenesis. VDR is an intracellular hormone receptor expressed in both normal epithelial and cancer colon cells at various levels. Several VDR gene polymorphisms, including FokI and BsmI, have been evaluated for their possible association with CRC susceptibility. The aim of our study was to investigate these two SNPs for the first time in Romanian CRC patients. FokI (rs228570 C>T) and BsmI (rs1544410 A>G) were genotyped by real-time polymerase chain reaction (RT-PCR) in 384-well plates using specific TaqMan predesigned probes on a ViiA™ 7 RT-PCR System. A total of 441 subjects (166 CRC patients and 275 healthy controls) were included. No statistically significant difference was observed between CRC patients and controls when we compared the wild-type genotype with heterozygous and mutant genotypes for both FokI (OR 0.85, 95% CI: 0.56-1.28; OR 0.95, 95% CI: 0.51-1.79, respectively) and BsmI (OR 0.97, 95% CI: 0.63-1.49; OR 1.10, 95% CI: 0.65-1.87, respectively) or in the dominant and recessive models. Also, we compared allele frequencies, and no correlation was found. Moreover, the association between these SNPs and the tumor site, TNM stage, and histological type was examined separately, and there was no statistically significant difference. In conclusion, our study did not show any association between FokI and BsmI SNPs and CRC susceptibility in a Romanian population. Further studies including a larger number of samples are needed to improve our knowledge regarding the influence of VDR polymorphism on CRC susceptibility.

DRFL: Dynamic-Recall Focal Loss for Image Classification and Segmentation

ABSTRACT The accuracy of neural networks heavily depends on numerous precisely annotated samples. But in actual datasets, the number of samples for each category varies greatly. In the dataset, some classes may have a large sample size, called majority classes, while others may have a small sample size, called minority classes. Training model with imbalanced data is often conducive to the majority classes, while unfair to minority classes. As a result, the outputs return good performance on majority classes and bad performance on minority classes. This article proposes a new loss function called dynamic-recall focal loss (DRFL), which can solve the problem of imbalanced data categories in image classification and medical segmentation tasks. The DRFL assigns different weight coefficient to the classes according to their dynamic recall based on focal loss. Experimental results have shown that the newly proposed loss function DRFL can effectively improve the classification and segmentation accuracy of two imbalanced datasets.

Computationally efficient and error aware surrogate construction for numerical solutions of subsurface flow through porous media

Limiting the injection rate to restrict the pressure below a threshold at a critical location can be an important goal of simulations that model the subsurface pressure between injection and extraction wells. The pressure is approximated by the solution of Darcy’s partial differential equation for a given permeability field. The subsurface permeability is modeled as a random field since it is known only up to statistical properties. This induces uncertainty in the computed pressure. Solving the partial differential equation for an ensemble of random permeability simulations enables estimating a probability distribution for the pressure at the critical location. These simulations are computationally expensive, and practitioners often need rapid online guidance for real-time pressure management. An ensemble of numerical partial differential equation solutions is used to construct a Gaussian process regression model that can quickly predict the pressure at the critical location as a function of the extraction rate and permeability realization. The Gaussian process surrogate analyzes the ensemble of numerical pressure solutions at the critical location as noisy observations of the true pressure solution, enabling robust inference using the conditional Gaussian process distribution.Our first novel contribution is to identify a sampling methodology for the random environment and matching kernel technology for which fitting the Gaussian process regression model scales as O(nlogn) instead of the typical O(n3) rate in the number of samples n used to fit the surrogate. The surrogate model allows almost instantaneous predictions for the pressure at the critical location as a function of the extraction rate and permeability realization. Our second contribution is a novel algorithm to calibrate the uncertainty in the surrogate model to the discrepancy between the true pressure solution of Darcy’s equation and the numerical solution. Although our method is derived for building a surrogate for the solution of Darcy’s equation with a random permeability field, the framework broadly applies to solutions of other partial differential equations with random coefficients.

Research on the energy absorption performance of lattice filling structures based on homogenization analysis technique

The lattice structure has great application prospects in the field of energy absorption, but the high cost of numerical analysis makes it difficult to carry out related structural designs. Therefore, this article studies the feasibility of the homogenization analysis technique for the analysis of energy absorption processes in lattice-filled structures. In the study, compression experiments and numerical simulations were conducted for four different types of unit cells (including periodic truss-lattice cells and triply periodic minimal surface (TPMS cells). In the experiments, two compression samples with different numbers of periods were designed for the same unit cell structure. The research shows that the homogenized numerical model takes a lot less time to run than the classical beam and shell element models. The results of the simulation are also in line with the data from the experiments. On the other hand, the material curve constructed based on a small number of unit cell samples has excellent analytical accuracy. This means that relevant research can obtain the material experimental data required for homogenized analysis using lattice samples with low manufacturing costs. Understanding these phenomena can provide a basis for the design of energy absorption structures.

The Effect Of Price And Content Marketing On Skintific Product Purchase Decisions In Medan Sunggal District, North Sumatra

The aim of this research is to determine the influence of price and content marketing on purchasing decisions for Skintific products. This research uses quantitative methods. The population in this study were in Medan Sunggal Sub-district, North Sumatra. The number of samples in this research was 100 samples. Data was collected by distributing questionnaires to respondents using Google Form. The results of data processing show that the price variable has a positive and significant influence on purchasing decisions for Skintific products. The content marketing variable has a positive and significant influence on purchasing decisions for Skintific products. Simultaneous research also shows that price and content marketing also have a positive and significant influence on purchasing decisions for Skintific products

Base editing screens define the genetic landscape of cancer drug resistance mechanisms.

Drug resistance is a principal limitation to the long-term efficacy of cancer therapies. Cancer genome sequencing can retrospectively delineate the genetic basis of drug resistance, but this requires large numbers of post-treatment samples to nominate causal variants. Here we prospectively identify genetic mechanisms of resistance to ten oncology drugs from CRISPR base editing mutagenesis screens in four cancer cell lines using a guide RNA library predicted to install 32,476 variants in 11 cancer genes. We identify four functional classes of protein variants modulating drug sensitivity and use single-cell transcriptomics to reveal how these variants operate through distinct mechanisms, including eliciting a drug-addicted cell state. We identify variants that can be targeted with alternative inhibitors to overcome resistance and functionally validate an epidermal growth factor receptor (EGFR) variant that sensitizes lung cancer cells to EGFR inhibitors. Our variant-to-function map has implications for patient stratification, therapy combinations and drug scheduling in cancer treatment.

Improved MRF rail surface defect segmentation method based on clustering features

Aiming at the characteristics of small number and many types of rail surface defect samples, as well as the problems of unstable transfer learning effect and threshold segmentation being easily affected by environmental factors in real scenes, an improved Markov defect segmentation method with zero samples is proposed. Firstly, the collected data is processed by Gabor function to highlight the defect features and reduce the data dimension to obtain the reduced dimension feature map; Kmeans clustering is performed on the processed feature map to reduce the distribution of data and reduce the influence of reflection and shadow, and the clustering result is used as the pre-classification matrix; an improved Markov random field two-layer graph model is constructed and inferred through the reduced dimension feature map and the pre-classification matrix; the local geometric structure of the defect part is analyzed according to the eigenvalues of the classification matrix inferred by the model; finally, the defect area is marked and the defect segmentation is completed. The experimental part uses a self-sampling data set, and the final conclusion is drawn based on the comparative experiment and ablation experiment. The experimental results show that the pixel accuracy, average pixel accuracy, weighted intersection-over-union ratio, and average intersection-over-union ratio of this method on the self-sampling data set are respectively 93.6%、80.7%、89.4%、68.2% , which exceeds the accuracy of other comparative detection algorithms.

A new omnibus SPRT chart for monitoring process mean and variability based on the average number of observations to signal

The recent development of the omnibus sequential probability ratio test (OSPRT) chart marks a significant contribution to the advancement of joint monitoring schemes. As the OSPRT chart is a variable-sample-size control chart, practitioners often wish to understand its inspection efficiency, i.e. the number of observations it samples before producing a signal. In this article, we propose two enhanced optimization designs for the OSPRT chart based on the average number of observations of signal (ANOS) and expected value of the ANOS (EANOS) metrics under deterministic and unknown shift sizes, respectively. The ANOS metric is central to our design as it perfectly combines both the average run length (ARL) and the average sample number. A comparative analysis reveals that the OSPRT chart outperforms four benchmarking control charts in terms of the ANOS and EANOS metrics. Finally, an implementation of the OSPRT chart is presented with a ball shear test dataset.

Lactose Breath Test: Possible Strategies to Optimize Test Performance, Accuracy, and Clinical Impact.

Lactose malabsorption (LM) refers to the incomplete absorption of lactose in the small intestine, resulting in the arrival of ingested lactose in the colon, which can give rise to symptoms defined as lactose intolerance (LI). The lactose breath test (LBT), thanks to its low cost, availability, and noninvasiveness, is the most used diagnostic method. However, the LBT is a tedious tool, requiring prolonged involvement of patients, qualified staff, and infrastructure, of which the most time-consuming factor is the frequency and number of breath samples needed. Objectives: To simplify the current LBT methodology, compliant with the current guidelines' statements, by reducing the test duration or the number of breath samples, without compromising the test's accuracy. Methods: The results of the standard LBT were compared with two simplified tests: a "shortened" test, lasting three hours, with samples taken every 30 min; and a "five-sample" test, lasting four hours, with samples taken every hour. Patients were stratified into three grades of malabsorption (mild, moderate, severe) based on the amount of gas exhaled. A clinical severity score was introduced to assess the clinical relevance of LI using a specific questionnaire. Results: Among the 543 patients enrolled (F 71.5%, mean age 43.7 ± 17.6 yrs), 60.4% (328/543) tested positive for LM. A total of 70.5% (383/543) presented LI, with 32.1% of those being true intolerants (LI without LM). The shortened test demonstrated an accuracy of 93.9%, with a sensitivity of 89.9% and a false negative rate of 10.1% (33/328). The five-sample test showed higher accuracy and sensitivity than the shortened test (96.5% and 94.2%, respectively; p = 0.03) with a false negative rate of 5.8% (19/328). Of the 19 false negatives in the five-sample test, 95% (18/19) were categorized as mild malabsorbents. No statistical correlation was found between the clinical severity score and LBT results. Conclusions: The five-sample test, involving hourly breath measurements, is a reliable option for simplifying the LBT without significantly reducing the procedure's sensitivity.

P27.22.B SEMISITTING POSITION VS. STANDARD SUPINE POSITION IN REDUCING INTRACEREBRAL BLEEDINGS DURING STEREOTACTIC BRAIN BIOPSY

Abstract BACKGROUND Stereotactic brain biopsy (SBB) represents a minimally invasive procedure to acquire tissue samples for histological diagnosis. Despite SBB’s many benefits, there is a chance of hemorrhagic complications. The incidence of hemorrhagic complications following SBB is not well established, but studies suggest that it may be higher than previously thought. In semi-sitting position (30° elevation of the head), the intracranial pressure is lower than in supine position due to decreased venous congestion. In our hypothesis, intraoperative bleeding may be reduced by exploiting semi-sitting position, potentially leading to decreased surgical morbidity. MATERIAL AND METHODS We conducted a retrospective monocentric 2-arms cohort study at our institution (Istituto “C. Besta” in Milan, Italy). We reviewed all patients who underwent stereotactic brain biopsy from the 1st of March 2023 to the 1st of April 2024. The primary endpoint is to demonstrate the improved safety of the semi-sitting position, as assessed by the absence of intracranial bleeding on postoperative CT scans compared to standard supine position. RESULTS Our analysis covered 78 patients, with 39 undergoing biopsy in the semi-sitting position and the remaining 39 in supine positions. Out of those in the semi-sitting position, 5 exhibited a hemorrhagic spot along the catheter’s edge or within the lesion as revealed by CT scans. This contrasts with the 14 instances found among patients not in the semi-sitting position, indicating a statistically significant difference (p-value: 0.028). However, no other significant correlations were observed in CT findings, such as bleeding, when compared to age (p-value: 0.352), biopsy site (p-value: 0.355), number of samples (p-value: 0.141), or histology (p-value: 0.251). CONCLUSION This study can potentially shed light on an innovative approach for patient positioning in stereotactic brain biopsy. The observed decrease in intraoperative bleeding may decrease procedure-related risks and surgical morbidity. As a result, patient safety and outcomes can be improved.

The Effect of Using Genre-Based Learning Method on Students’ Writing Descriptive Text at SMPN 1 Paranginan

This research aims to determine the effect of using the Genre-based learning method on students' writing ability. This research was prepared using quantitative experimental research with a sample of class IX students at SMP N 1 Paranginan. The number of samples in this study was 40 students and was divided into two classes, namely the experimental class and the control class. The data collection instrument uses a writing test and the data collection technique uses a pre-test and post-test, while data was analyzed by using T-test Formula. From the research that has been conducted, it was found that there was a significant difference between the pre-test and post-test. In the experimental class, the score was 53,05 and the post-test score was 81.4, while in the control class, the pre-test score was 56,1 and the post-test score was 80.95. So, it can be concluded that the use of the genre-based learning method has a great influence on improving students' writing skills at SMP N 1 Paranginan.

Data-driven machine learning approaches for precise lithofacies identification in complex geological environments

ABSTRACT Reservoir characterization is a vital task within the oil and gas industry, with the identification of lithofacies in subsurface formations being a fundamental aspect of this process. However, lithofacies identification in complex geological environments with high dimensions, such as the Lower Indus Basin in Pakistan, poses a notable challenge, especially when dealing with limited data. To address this issue, we propose four common data-driven machine learning approaches: multi-resolution graph-based clustering (MRGC), artificial neural networks (ANN), K-nearest neighbors (KNN), and self-organizing map (SOM). We utilized these proposed approaches to assess their performance in scenarios with varying core sample availability, specifically evaluating their effectiveness in identifying lithofacies within the Lower Goru formation of the middle Indus Basin. The study reveals that in scenarios with a limited number of core samples, MRGC is the preferred choice, while KNN or MRGC is more suitable for larger datasets. The results demonstrate the superior performance of MRGC and KNN in lithofacies identification within the specified geological environment, with SOM following closely behind, and ANN exhibiting comparatively lower efficacy. The accurate identification of lithofacies from the selected model is complemented by the application of the truncated Gaussian simulation method for facies modeling. Comparative results confirm the excellent agreement between the model identification of lithofacies from well logs and electro-facies obtained from the truncated Gaussian simulation electro-facies volume. This study highlights the crucial role of selecting the right machine learning approach for precise lithofacies identification and modeling in complex geological environments. The comparative analysis provides practitioners in the petroleum industry with insights into the strengths and limitations of each method, enhancing existing knowledge. In conclusion, this research emphasizes the significance of comprehensive research and method selection for advancing lithofacies identification in diverse formations or study areas, ultimately benefiting the broader field of subsurface characterization in the petroleum industry.

The incident and Molecular Identification of the Cryptosporidium Parvum in Cultivated Carp Fish (Cyprinus carpio) in Baghdad

The current study aimed to investigate the occurrence of Cryptosporidium species by traditional and molecular methods in freshwater fish in Baghdad city. A convenient number of stomach and intestinal continent samples were collected from three types of freshwater fish from different areas of Baghdad city from November, 2023 to April, 2024. All samples were examined phenotypically using a modified Ziehl-Neelsen stain and genotypically (nested PCR technique) based on a partial sequence of 18SrRNA genes with sequencing and phylogenetic tree analysis. The overall infection rate of fish with C. parvum was 5.5% (10/180). The common carp (Cyprinus carpio) was the only fish type infected, with a rate of 16.66% (10 60). January and February had the highest infection rate, 3.33%. The molecular identification and sequencing confirmed the infections with C. parvum. Molecular speciation of the parasite confirmed microscopical primary identification results. This study suggests that fish became a new source for transmission to humans and livestock.

Regulate the Risk Factors of Pseudomonas aeruginosa by some of Sigma Factor Genes

In this research, a study was conducted isolating and identification of Pseudomonas aeruginosa bacteria from clinical samples from various pathological conditions such as burns , otitis media , urinary tract infection , wounds, eye infection, blood infection, catheter infection, throat , vaginal infection and surgical. The present study includes 313 samples collected from patients who attend Alsader Medical City and Al- Hakim General Hospital during the period extended From September 2023 to December 2023 ( men and women ) with age groups between (1 -77) years. The identification of bacterial isolates were detected by classical and molecular technique (PCR), were the frequency among males 163 (52.07% ) was more than that in female 150 (47.92). The samples distribution is made according age group , it appears high for the following high (15.97) and (12.77) with group (16-30) years of male and female respectively . The results revealed that 57/313 specimens of the total number of samples are Pseudomonas aeruginosa by identified according to their morphological, physiological and biochemical properties. Also with molecular identification by using four primers wre the results showed 16S rRNA gene was detected in 57/57 (100%) P. aeruginosa isolates , the Las biofilm gene was detected in 43/57 (75.43%) , the algD gene was detected in 41/57 (71.92%) and 41/57 samples show positive results with the oprL gene. Six types of antibiotics were also tested to determine the resistance of the samples to them.

SALIVA FLOW RATE AND PLAQUE SCORE IN CONVENTIONAL SMOKERS

Background: Smoking had become a common habit in the community. The Lampung Dormitory in Yogyakarta had found that people who had consumed cigarettes for more than 1 year had a moderate level of dental and oral hygiene. Teeth and mouths could be said to be clean if they were free of dirt such as debris, plaque, and calculus. Purpose: This study aimed to analyze the relationship between saliva flow rate and plaque score in conventional smokers in Lampung Dormitory. Methods: A type quantitative research with an observational analytical approach was used, employing a cross-sectional design. The population of this study was conventional cigarette users aged 18–30. The number of samples in this study was 70 respondents with sampling using the purposive sampling technique. The research variables were saliva flow rate and plaque score. Saliva was taken by the spitting method, and the plaque scores were measured using the PHP (Personal Hygiene Performance) index. Data analysis used bivariate analysis in the form of a chi-square test. Results: Smokers had a low criterion saliva flow rate (50.0%), and smokers had a poor criterion plaque score (45.7%). The average saliva flow rate was low (0.29±0.08), and the plaque score was poor (2.85±0.81). based on the chi-square test showed that p-value was 0.000. Conclusion: There was a significant relationship between saliva flow rate and plaque score in conventional smokers.