27 research outputs found
Growth status and physiological changes of sugar beet seedlings in response to acidic pH environments
No full textSugar beet (Beta vulgaris L.) is an important sugar crop that is popularly cultivated in a variety of agriculture conditions. Here, we studied sugar beet growth in different pH soils (pH 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, and 9.0) and analyzed their growth status and physiology. Sugar beet growth was best at pH 9.0 and worst at pH 5.0. As the soil pH decreased from 9.0 to 5.0, the osmoregulatory substances, antioxidant enzyme activity, and elemental contents in leaves and roots showed increasing trends, while photosynthesis and macronutrient contents showed decreasing trends. To explore the physiological mechanisms sugar beet use to respond to different pH environments, we analyzed the correlations between leaf net photosynthesis rate and physiological changes and nutrient contents of sugar beet. One of the factors inhibiting sugar beet growth in low pH soils was a reduction in photosynthetic capacity. The accumulation of osmoregulatory substances and increased peroxidative damage may have led to the decrease in leaf net photosynthesis rate. Furthermore, the decrease in nutrient content and accumulation of metal elements were correlated with the decrease in leaf photosynthetic rate. QRT-PCR analysis showed higher expression levels of antioxidant enzyme genes in the leaves and roots of sugar beet grown in low pH environments compared to those in high pH environments. Correspondingly, antioxidant enzyme activity was significantly higher in beets in low pH environments than in beets in high pH environments. These results provide important insight into the physiological responses by which sugar beet can adapt to different pH soils
An All-In-One NLP Stock Market Backtester:
No full textDespite the popularity, we noticed that it is rather hard to verify a NLP/text-mining like stock prediction model's performance due to the amount of "groundwork" needed. It is very typical a researcher will have to gather the plain text data, the company info, the stock market data, and categorize them in a way that is communicable with each other and the model; then the researcher will need to build a virtual trading platform that keeps track of all the trading signals generated by the model, log the activities in a certain way, then do some kinds of visualization for evaluations. To implement all these steps from ground up, it is required for a researcher to have certain level of proficiency on skills which are, from a research stand-point, fairly deviated from the nature of the NLP/text-mining model itself (like scraping a website and understanding the fundamental mechanism of trading in stock market). Thus, we like to build a set of lightweight tools that may automate such process to a certain degree
Moderate deviations for a class of recursions
No full textAbstractIn this paper, we establish a moderate deviation principle for a class of recursions which have the form of Zn+1=(1−Γn+1gn)Zn+Vn+1gn, where gn are constants, Vn,Γn are random variables for any n≥1. These recursions often occur in stochastic approximation algorithms
Design and implementation of commodity information query system based on Android and mobile Internet
No full textInterfacial Stresses of Thermal Barrier Coating with Film Cooling Holes Induced by CMAS Infiltration
No full textTo obtain high gas turbine efficiency, a film cooling hole is introduced to prevent the destruction of thermal barrier coating systems (TBCs) due to hot gases. Furthermore, environmental calcium-magnesium-aluminum-silicate (CMAS) particulates plug the film cooling hole and infiltrate the TBCs to form a CMAS-rich layer, which results in phase transformations and significant modifications in the thermomechanical properties that impact the TBCs during cooling. This study aimed to establish a three-dimensional thermo-fluid-solid coupling TBCs model with film cooling holes and CMAS infiltration to analyze the temperature and residual stress distribution via simulations. For the interfacial stress around the cooling hole at the TC/BC interface, the film cooling holes alleviated the interfacial residual stress by 60% due to the reduction in temperature by 40%. In addition, CMAS infiltration intensified the interfacial residual stress via phase transformation. As a result of the influence of larger penetration depths and expansion rates of phase transformation, a significant increase in residual stress was observed. At the beginning of CMAS infiltration, the interfacial stress would be more dominated by the effect of infiltration depth. In addition, the failure due to interfacial normal and tangential stresses was more likely to be found at the infiltration zone near the cooling hole
Unsupervised Hashing with Gradient Attention
No full textThe existing learning-based unsupervised hashing method usually uses a pre-trained network to extract features, and then uses the extracted feature vectors to construct a similarity matrix which guides the generation of hash codes through gradient descent. Existing research shows that the algorithm based on gradient descent will cause the hash codes of the paired images to be updated toward each other’s position during the training process. For unsupervised training, this situation will cause large fluctuations in the hash code during training and limit the learning efficiency of the hash code. In this paper, we propose a method named Deep Unsupervised Hashing with Gradient Attention (UHGA) to solve this problem. UHGA mainly includes the following contents: (1) use pre-trained network models to extract image features; (2) calculate the cosine distance of the corresponding features of the pair of images, and construct a similarity matrix through the cosine distance to guide the generation of hash codes; (3) a gradient attention mechanism is added during the training of the hash code to pay attention to the gradient. Experiments on two existing public datasets show that our proposed method can obtain more discriminating hash codes
HyperSMamba: A Lightweight Mamba for Efficient Hyperspectral Image Classification
No full textDeep learning has recently achieved remarkable progress in hyperspectral image (HSI) classification. Among these advancements, the Transformer-based models have gained considerable attention due to their ability to establish long-range dependencies. However, the quadratic computational complexity of the self-attention mechanism limits its application in hyperspectral image classification (HSIC). Recently, the Mamba architecture has shown outstanding performance in 1D sequence modeling tasks owing to its lightweight linear sequence operations and efficient parallel scanning capabilities. Nevertheless, its application in HSI classification still faces challenges. Most existing Mamba-based approaches adopt various selective scanning strategies for HSI serialization, ensuring the adjacency of scanning sequences to enhance spatial continuity. However, these methods lead to substantially increased computational overhead. To overcome these challenges, this study proposes the Hyperspectral Spatial Mamba (HyperSMamba) model for HSIC, aiming to reduce computational complexity while improving classification performance. The suggested framework consists of the following key components: (1) a Multi-Scale Spatial Mamba (MS-Mamba) encoder, which refines the state-space model (SSM) computation by incorporating a Multi-Scale State Fusion Module (MSFM) after the state transition equations of original SSMs. This module aggregates adjacent state representations to reinforce spatial dependencies among local features; (2) our proposed Adaptive Fusion Attention Module (AFAttention) to dynamically fuse bidirectional Mamba outputs for optimizing feature representation. Experiments were performed on three HSI datasets, and the findings demonstrate that HyperSMamba attains overall accuracy of 94.86%, 97.72%, and 97.38% on the Indian Pines, Pavia University, and Salinas datasets, while maintaining low computational complexity. These results confirm the model’s effectiveness and potential for practical application in HSIC tasks
Investigation of region-of-interest-based functional connectivity within the default mode network among adolescents with depression complicated by obesity
No full textAbstract Background This study utilized resting-state functional magnetic resonance imaging (rs-fMRI) combined with default mode network (DMN) region-of-interest (ROI)-based functional connectivity (FC) analysis to examine adolescents with depression complicated by obesity. Correlation analyses were conducted between the functional connectivity of distinct brain regions in the depression-with-obesity cohort and scores from the Adolescent Self-Rating Life Events Checklist (ASLEC) to examine potential associations. This study aimed to elucidate the underlying pathogenesis of depression complicated by obesity and identify potential imaging biomarkers for early diagnosis in this population. Methods We analysed rs-fMRI data from 37 adolescents with depression complicated by obesity (OMDD group, n = 37), 38 patients with depression (MDD group, n = 38), and 35 healthy controls (HCs group, n = 35). DMN FC was compared with whole-brain connectivity during rs-fMRI. Imaging data from the three groups were collected and analysed using one-way analysis of variance (ANOVA). Group differences in FC values were assessed, and correlations were examined between these values and clinical scale scores across patient groups. Results Compared with the MDD group, the OMDD group demonstrated significantly increased FC between the left parahippocampal gyrus (left PHG) and the right precuneus. Compared with the HCs, both the OMDD and MDD patients presented reduced FC between the left PHG and the left putamen, right putamen, and opercular part of the right inferior frontal gyrus. All findings were corrected using Gaussian random field (GRF) theory at a voxel-level threshold of P 30 voxels). Analyses revealed significant negative correlations between the FC values of the right putamen in the OMDD group and the “interpersonal relationship” scores (r = − 0.373, P = 0.023). Conclusion Compared with healthy controls, adolescents with depression complicated by obesity demonstrated significant alterations in functional integration within the DMN. Specifically, the OMDD group exhibited aberrant FC between the left PHG and the right precuneus. This aberrant FC may underlie the pathophysiology of this disorder. These findings offer novel insight into the neural mechanisms of depression comorbid with obesity, enhancing our understanding of its pathogenesis
Role of syndecan-1 and exogenous heparin in hepatoma sphere formation
No full textGlycosaminoglycan-modified proteoglycans play important roles in many cell activities, including cell differentiation and stem cell development. Tumor sphere formation ability is one of properties in cancer stem cells (CSCs). The correlation between CSC markers and proteoglycan remains to be clarified. Upon hepatoma sphere formation, expression of CSC markers CD13, CD90, CD133, and CD44, as well the syndecan family protein syndecan-1 (SDC1), increased as analyzed by PCR. Further examination by suppression of CD13 expression showed downregulation of SDC1 and CD44 gene expression, whereas suppression of SDC1 gene expression downregulated CD13 and CD44 gene expression. Suppression of SDC1 gene expression also suppressed sphere development, as analyzed by a novel sphereocrit assay to quantify the level of sphere formation. The heparin disaccharide components, but not those of chondroitin disaccharide, changed with hepatoma sphere development, revealing the increased levels of N-sulfation and 2-O-sulfation. These explained the inhibition of hepatoma sphere formation by exogenous heparin. In conclusion, we found that SDC1 affected CSC marker CD13 and CD44 expression. SDC1 proteoglycan and heparin components changed and affected hepatoma sphere development. Application of heparin mimics in reduction of hepatoma stem cells might be possible.The accepted manuscript in pdf format is listed with the files at the bottom of this page. The presentation of the authors' names and (or) special characters in the title of the manuscript may differ slightly between what is listed on this page and what is listed in the pdf file of the accepted manuscript; that in the pdf file of the accepted manuscript is what was submitted by the author
High-Throughput Screening of Metal–Organic Frameworks Assisted by Machine Learning: Propane/Propylene Separation
No full textThe separation of a propane (C3H8)/propylene(C3H6) mixture
is of paramount importance in the petrochemical
industry. Metal–organic frameworks (MOFs), as a class of promising
alternative to the traditional adsorbents, have garnered extensive
interest. This study proposes a machine learning-assisted high-throughput
screening strategy for the identification of suitable MOFs for C3H8/C3H6 separation, striving
to accelerate the discovery of high-performance MOF candidates for
this particular application. First, a chemical/geometric analysis-based
prescreening is applied to a data set of 146 203 MOFs composed
of an experimentally synthesized MOF database and a hypothetical MOF
database, and MOFs with undesirable chemical/geometric features were
excluded. Six structural and nine chemical descriptors were calculated
for the remaining MOFs. Random Forest regression algorithm was applied
to “learn” the relationship correlations between the
features (chemical and/or structural) of MOFs and their C3H8/C3H6 separation capacity. Grand
Canonical Monte Carlo (GCMC) simulations were applied to evaluate
the C3H8/C3H6 separation
performances of the randomly selected training and testing MOF samples.
A performance prediction model based on chemical and structural descriptors
was obtained with R2 equal to 0.96, which
was employed for a separation performance prediction of the remaining
MOFs. 2500 MOFs with potential to possess high C3H8/C3H6 separation performance were shortlisted
by the prediction model. GCMC simulations were applied to calibrate
the prediction results and validate of the machine learning model.
MOFs with competitively high C3H8/C3H6 separation potential and regenerability were identified,
and a comparison with MOFs reported in the literature was made, indicating
that the proposed machine learning-assisted high-throughput screening
approach is efficient and effective. Furthermore, structure–property
correlation analysis was conducted
