Abstract Background Microbial adaptation to salinity has been a classic inquiry in the field of microbiology. It demonstrated that microorganisms can endure stress via either “salt-in” strategy, involving inorganic ion uptake, or “salt-out” relying on compatible solutes. While these insights are mostly based laboratory-cultured isolates, exploring adaptive mechanisms within natural gradient is crucial for gaining deeper understanding microbial estuarine ecosystem. Results Here, we conducted...

Surface defect inspection is a key step to ensure the quality of hot rolled steel surface. However, current advanced detection (DET) methods have high precision but low speed, which hinders application detector in actual production. In this work, real-time network (RDN) focusing on both speed and accuracy proposed solve problem surface detection. RDN takes ResNet-dcn, modular encoding, decoding with light weights, as basic convolutional architecture whose backbone pretrained ImageNet. To...

Recently, vision-language models have made remarkable progress, demonstrating outstanding capabilities in various tasks such as image captioning and video understanding. We introduce Valley2, a novel multimodal large language model designed to enhance performance across all domains extend the boundaries of practical applications e-commerce short scenarios. Notably, Valley2 achieves state-of-the-art (SOTA) on benchmarks, surpassing open-source similar size by margin (79.66 vs. 72.76)....

ABSTRACT Background Immunotherapy, especially immune checkpoint blockade (ICB) therapy, has demonstrated noteworthy advancements in the realm of non‐small cell lung cancer (NSCLC). However, efficacy ICB therapy is limited to a small subset patients with NSCLC, and underlying mechanisms remain poorly understood. Study Design Discoveries In this study, we conducted comprehensive investigation metabolic profiles infiltrating T cells NSCLC tumors revealed heterogeneity, which associated...

Surface defect classification of hot-rolled strip based on machine vision is a challenge task caused by the diversity morphology, high inter-class similarity, and real-time requirements in actual production. In this work, VGG16-ADB, an improved VGG16 convolution neural network, proposed to address problem identification strip. The network takes as benchmark model, reduces system consumption memory occupation reducing depth width structure, adds batch normalization layer accelerate...

State estimators are crucial for the effective use of batteries in real-world applications. Insufficient algorithms can lead to user dissatisfaction, safety risks, and accelerated battery degradation, posing significant risks manufacturers. Developing management systems (BMS) involves defining requirements, implementing algorithms, validating them, which is a complex process. The performance BMS influenced by constraints related hardware, data storage, calibration processes during...

The computational methods of protein-protein interaction sites prediction can effectively avoid the shortcomings high cost and time in traditional experimental approaches. However, serious class imbalance between interface non-interface residues on protein sequences limits performance these methods. This work therefore proposed a new strategy, NearMiss-based under-sampling for unbalancing datasets Random Forest classification (NM-RF), to predict sites. Herein, were represented by...

Predicting traffic speed accurately is a very challenging task of the intelligent system (ITS), due to complex and dynamic spatial-temporal dependencies from both temporal spatial aspects. There not only exits short-term local neighboring fluctuation long-term global trend in aspect, but also correlations aspect. Most existing work focus on dependencies, ignoring corrections, which comparably critical for prediction. To address this problem, we propose novel Dynamic Global-Local...

Surface defect classification plays an important role in the assessment of production status and analyzing possible causes hot rolled strip steel. It is extremely challenging owing to rare occurrence various appearances defects. In this work, improved deep learning model proposed solve problem poor accuracy when only a few labeled samples can be available. Different from most inductive small-sample methods, transductive algorithm designed where new classifier trained test phase therefore fit...

Bearings play an important role as the connection between motor and gear. At present, data collected by most bearing datasets are vibration signals in one-dimensional time domain, then convolution or other methods used to analyze signals. In this work, a fault diagnosis method based on continuous wavelet transform scalogram (CWTS) multi-scale convolutional neural network (MS-CNN) is proposed paper. Continuous time-frequency relationship of signal extract frequency information signal, two...

K-means clustering algorithm is one of the most popular technique for in machine learning, however, existing k-means algorithm, ability different features and importance data objects are treated equally; discriminative cannot be differentiated effectively. In light this limitation, paper put forward an enhanced regularized type with adaptive weights which we introduced feature matrix vector into objective function developed a new l2-norm regularization to features, then obtained...

Purpose The purpose of this paper is to solve the problem existing in forecast impact disturbance grey system. Design/methodology/approach Under axiomatic system buffer operator theory, a novel kind operators with variable weight λ based on principle average tempo time sequence and using new information proposed. optimization solution for obtained by genetic algorithm. It proved that are effective. Findings results show accord operator's three axioms monotonicity non‐variable axiom....

Fuzzy c-means algorithm (Fcm) frequently applid in machine learning has been proven an effective clustering approach. However, the traditional Fcm cannot distinguish importance of different data objects and discriminative ability features process. In t his paper, we propose a new kind framework: DwfwFcm.Considering weights feature weights, adaptive vector matrix are introduced into conventional objective function is constructed. By proposed function, corresponding scientific updating...

The influence of morphological defects on long-term reliability 1200-V/40-A 4H-SiC junction barrier Schottky (JBS) diodes under the 668 h reverse bias stress (RBS) is investigated in depth. variation electrical properties and related degradation mechanisms analyzed. It found that breakdown voltage ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\textit{V}_{\text{BR}}$</tex-math> </inline-formula> )...

Abstract The purpose of clustering is to partition data similar with each other into a same group and dissimilar different groups. However, in most existing fuzzy approaches, the membership degrees an individual belonging clusters are relied on distances between cluster centroids, similarity ignored, besides, outliers cannot be distinguished effectively. For improving performance addressing problems based concept that close should grouped together assigned cluster, this paper, we present...