Explore the vast range of titles available.

Skin diseases are common conditions that pose a significant threat to human health, and automated classification plays an important role in assisting clinical diagnosis. However, existing image classification approaches based on convolutional neural networks (CNNs) and Transformers have inherent limitations. CNNs are constrained in capturing global features, whereas Transformers are less effective in modeling local details. Given the characteristics of dermoscopic images, both local and global features are equally crucial for classification tasks. To address this issue, we propose an improved Swin Transformer-based model, termed MaLafFormer, which incorporates a Modulated Fusion of Multi-scale Attention (MFMA) module and a Lesion-Area Focus (LAF) module to enhance global modeling, emphasize critical local regions, and improve lesion boundary perception. Experimental results on the ISIC2018 dataset show that MaLafFormer achieves 84.35% ± 0.56% accuracy (mean of three runs), outperforming the baseline 77.98% ± 0.34% by 6.37%, and surpasses other compared methods across multiple metrics, thereby validating its effectiveness for skin lesion classification tasks.

The stock market is characterized by its high nonlinearity and complexity, making traditional methods ineffective in capturing its nonlinear features and complex market dynamics. This paper proposes a novel stock price forecasting model—the Variational Mode Decomposition—Triangulated Maximally Filtered Graph—Long Short-Term Memory (VMD–TMFG–LSTM) combined model—aimed at improving prediction accuracy, stability, and computational efficiency. The proposed model first employs Variational Mode Decomposition (VMD) to decompose the stock price time series into multiple smooth intrinsic mode functions (IMFs), reducing data complexity and mitigating noise interference. Subsequently, the TMFG algorithm is utilized for feature selection, simplifying the input data and accelerating the iterative convergence process. Finally, the filtered features are modeled and predicted using a Long Short-Term Memory (LSTM) network. Experimental results demonstrate that the VMD–TMFG–LSTM model significantly outperforms AutoRegressive Integrated Moving Average (ARIMA), Neural Network (NN), Deep Neural Network (DNN), Convolutional Neural Network (CNN), as well as single LSTM, TMFG–LSTM, and VMD–LSTM models in forecasting the closing prices of multiple stocks. Specifically, for Shanghai International Airport Co., Ltd. (sh600009), the VMD–TMFG–LSTM model achieves a 69.76% reduction in Root Mean Squared Error (RMSE), a 71.41% reduction in Mean Absolute Error (MAE), a 46.28% reduction in runtime, and an improvement of 0.2184 in R-squared (R 2 ), indicating significantly higher prediction accuracy. In conclusion, the combined model proposed in this paper enhances the accuracy, efficiency, and stability of stock price prediction, providing a robust and efficient solution for forecasting stock market trends.

Medical image segmentation plays a crucial role in computer-aided diagnosis. By segmenting pathological tissues in medical images, doctors can observe anatomical structures more clearly, thereby achieving more accurate disease diagnoses. However, existing medical image segmentation networks have issues such as insufficient capability to extract features from target areas, as well as high number of parameters and increased computational complexity. To address these issues, a lightweight Dual-Encoding Feature Interaction Fusion network (DEFIF-Net) is proposed in this paper for medical image segmentation. Firstly, in the encoding stage of DEFIF-Net, a global dependency fusion branch is introduced as an additional encoder to capture distant feature dependencies, whereby the neighboring and distant feature dependencies are effectively integrated by the newly designed feature interaction fusion convolution. Secondly, between the encoder and decoder, channel feature reconstruction modules (CFRMs) are used to enhance the feature representation of important channels. Additionally, a novel multi-branch ghost module (MBGM) is used in the bottleneck layer of the network to enhance its efficiency in capturing and retaining different types of feature information. Lastly, a novel residual feature enhancement (RFE) decoder is utilized to emphasize boundary features, thereby increasing the network’s sensitivity to lesion boundaries. The segmentation performance of the proposed DEFIF-Net network is evaluated in two different medical image segmentation tasks. The obtained experimental results demonstrate that, compared to state-of-the-art networks, DEFIF-Net exhibits superior segmentation performance on all three datasets used, while also having a lower parameter count and computational complexity.

The Chinese stock market, one of the largest and most dynamic emerging markets, is characterized by individual investor dominance and strong policy influence, resulting in high volatility and complex dynamics. These distinctive features pose substantial challenges for accurate forecasting. Existing models like RNNs, LSTMs, and Transformers often struggle with non-stationary data and long-term dependencies, limiting their forecasting effectiveness. This study proposes a hybrid forecasting framework integrating the Non-stationary Autoformer (NSAutoformer), LASSO feature selection, and financial sentiment analysis. LASSO selects key features from diverse structured variables, mitigating multicollinearity and enhancing interpretability. Sentiment indices are extracted from investor comments and news articles using an expanded Chinese financial sentiment dictionary, capturing psychological drivers of market behavior. Experimental evaluations on the Shanghai Stock Exchange Composite Index show that LASSO-NSAutoformer outperforms the NSAutoformer, reducing MAE by 8.75%. Additional multi-step forecasting and time-window analyses confirm the method’s effectiveness and stability. By integrating multi-source data, feature selection, and sentiment analysis, this framework offers a reliable forecasting approach for investors and researchers in complex financial environments.

Current research on futures price prediction focuses on the autocorrelation of historical prices, yet the resulting predictions often suffer from issues of inaccuracy and lag. This paper uses Chinese corn futures as the subject of study. First, we identify key influencing factors, such as Chinese soybean futures, U.S. soybean futures, and the U.S.-China exchange rate, that exhibit ‘predictive causality’ with corn futures prices through the Granger causality test. We then apply the sample convolution and interaction network (SCINet) to perform both single-step and multi-step predictions of futures prices. The experimental results show that incorporating key influencing factors significantly improves prediction accuracy. For instance, in the single-step prediction, combining historical prices with Chinese soybean futures prices reduces the MAE and RMSE values by 5.12% and 3.45%, respectively, compared to using historical prices alone. Furthermore, the SCINet model outperforms traditional models such as temporal convolutional networks (TCN), gated recurrent units (GRU), and long short-term memory (LSTM) networks when based solely on historical prices. This study validates the effectiveness of key influencing factors in forecasting Chinese corn futures prices and demonstrates the advantages of the SCINet model in futures price prediction. The findings provide valuable insights for optimising the agricultural futures market and enhancing the ability to predict price risks.

In this paper, the nutritional ingredient, aroma component, and texture of three kinds of hickories, including American hickory, Chinese Linan hickory, and Chinese Hunan hickory, were tested by instruments. The quality of different hickory varieties was analyzed at three levels by using the grey entropy correlation analysis, namely, the single nutrient composition analysis; nutritional composition and texture analysis; nutrient composition, texture, and aroma analysis. Through the analysis of nutritional composition, American hickory gets the highest score (80.6945), followed by Linan hickory (74.9987), and Hunan hickory has the lowest score (58.5925). Through the analysis of nutrition composition and texture, Linan hickory has the highest score (80.89), American hickory is the second (71.77), and Hunan hickory is last (61.62). Through the analysis of nutrition composition, texture and aroma, Linan hickory has the highest score (75.91), followed by American hickory (74.17), and Hunan hickory has the lowest score (64.20). Finally, the comprehensive evaluation of Linan hickory quality index score is the highest. The main factors contributing to the high score of Linan hickory include superior fatty acid spectrum, aminogram and higher initial chewing hardness, moderate crispness of secondary chewing, optimal palatability, and unique aroma components ((S)-2-methyl-1-butanol, 3-methyl-2-pentene, (+/−)-2-methylbutyric acid methyl ester ethyl butyrate, ethyl 2-methylbutyrate, methyl phthalate, decene, (1S)-(−)-β-pinene). The research results provide a basis for consumers to understand the quality differences of different hickories.

The colors and shapes of different flowers are similar, requiring an accurate and efficient multitarget recognition method to identify flower species, which can assist in agricultural automation for harvesting. This article utilizes deep neural networks and transfer learning methods, employing computer vision technology for flower recognition and classification. It compares the detection accuracy and speed of different algorithms to achieve real‐time multitarget detection of flowers. First, a lightweight target detection model based on YOLOv4 is proposed, replacing the original CSPDarknet53 with lightweight networks such as MobileNetV3 or CSPDarknet53_(t)iny. The results show that the lightweight target detection model achieves more than 30 frames per second and has a higher mAP value. Second, an image classification model based on strong‐supervised deep learning is established, using ShuffleNet networks and transfer learning techniques. During this process, 80 images generated by a generative adversarial network are added to enhance the model's generalization ability. The results indicate that, compared to training from scratch, transfer learning accelerates model convergence and improves prediction accuracy. The model under strong supervision classification demonstrates stronger generalization ability and robustness. This article proposes a lightweight YOLOv4‐based detection model using MobileNetV3 or CSPDarknet53_(t)iny, achieving 30+ FPS and higher mAP. It also presents a ShuffleNet‐based classification model with transfer learning and GAN‐augmented images, improving generalization and accuracy. The study includes dataset construction, lightweight detection, and strongly supervised classification.

Garlic ( Allium sativum L.) is a compelling horticultural crop with high culinary and therapeutic values. Commercial garlic varieties are male-sterile and propagated asexually from individual cloves or bulbils. Consequently, its main breeding strategy has been confined to the time-consuming and inefficient selection approach from the existing germplasm. Polyploidy, meanwhile, plays a prominent role in conferring plants various changes in morphological, physiological, and ecological properties. Artificial polyploidy induction has gained pivotal attention to generate new genotype for further crop improvement as a mutational breeding method. In our study, efficient and reliable in vitro induction protocols of autotetraploid garlic were established by applying different antimitotic agents based on high-frequency direct shoot organogenesis initiated from inflorescence explant. The explants were cultured on solid medium containing various concentrations of colchicine or oryzalin for different duration days. Afterward, the ploidy levels of regenerated plantlets with stable and distinguished characters were confirmed by flow cytometry and chromosome counting. The colchicine concentration at 0.2% (w/v) combined with culture duration for 20 days was most efficient (the autotetraploid induction rate was 21.8%) compared to the induction rate of 4.3% using oryzalin at 60 μmol L –1 for 20 days. No polymorphic bands were detected by simple sequence repeat analysis between tetraploid and diploid plantlets. The tetraploids exhibited a stable and remarkable dwarfness effect rarely reported in artificial polyploidization among wide range of phenotypic variations. There are both morphological and cytological changes including extremely reduced plant height, thickening and broadening of leaves, disappearance of pseudostem, density reduction, and augmented width of stomatal. Furthermore, the level of phytohormones, including, indole propionic acid, gibberellin, brassinolide, zeatin, dihydrozeatin, and methyl jasmonate, was significantly lower in tetraploids than those in diploid controls, except indole acetic acid and abscisic acid, which could partly explain the dwarfness in hormonal regulation aspect. Moreover, as the typical secondary metabolites of garlic, organosulfur compounds including allicin, diallyl disulfide, and diallyl trisulfide accumulated a higher content significantly in tetraploids. The obtained dwarf genotype of autotetraploid garlic could bring new perspectives for the artificial polyploids breeding and be implemented as a new germplasm to facilitate investigation into whole-genome doubling consequences.

Candlestick charts provide a visual representation of price trends and market sentiment, enabling investors to identify key trends, support, and resistance levels, thus improving the success rate of stock trading. The research presented in this paper aims to overcome the limitations of traditional candlestick pattern analysis, which is constrained by fixed pattern definitions, quantity limitations, and subjectivity in pattern recognition, thus improving its effectiveness in dynamic market environments. To address this, a two-layer clustering method based on a candlestick sequence simlarity matching model is proposed for identifying valid candlestick patterns and constructing a pattern library. First, the candlestick sequence similarity matching model is used to address the pattern matching issue; then, a two-layer clustering method based on the K-means algorithm is designed to identify valid candlestick patterns. Finally, a valid candlestick pattern library is built, and the predictive ability and profitability of some patterns in the library are evaluated. In this study, ten stocks from different industries and of various sizes listed on the Shanghai Stock Exchange were selected, using nearly 1000 days of their data as the test set. The predictive ability of some patterns in the library was evaluated using out-of-sample data from the same period. This selection method ensures the diversity of the dataset. The experimental results show that the proposed method can effectively distinguish between bullish and bearish patterns, breaking through the limitations of traditional candlestick pattern classification methods that rely on predefined patterns. By clearly distinguishing these two patterns, it provides clear buy and sell signals for investors, significantly improving the reliability and profitability of trading strategies.

Salinity stress poses a major challenge to agricultural productivity worldwide, including for pumpkin, a globally cultivated vegetable crop with great economic value. To deal with salt stress, plants exhibit an array of responses such as changes in their root system architecture. However, the root phenotype and gene expression of pumpkin in response to different concentrations of NaCl remains unclear. To this end, this study evaluated the effects of salinity stress on root architecture in C. moschata (Cmo-1, Cmo-2 and Cmo-3) and C. maxima (Cma-1, Cma-2 and Cma-3), as well as their hybrids of C. moschata and C. maxima (Ch-1, Ch-2 and Ch-3) at the germination and seedling stages. The results showed that the total root length and the number of root tips decreased by more than 10% and 5%, respectively, under 180 mM NaCl conditions compared to those under the 0 mM NaCl conditions. In contrast, the total root length and the number of root tips were increased or decreased under 60 mM NaCl conditions. Meanwhile, salt stress was considered severe when treated with more than 120 mM NaCl, which could be used to evaluate the salt tolerance of the germplasm resources of pumpkin. In addition, the transcriptional changes in the roots of both Cmo-3 and Cma-2 under salt stress were analyzed via RNA-sequencing. We found 4299 and 2141 differential expression genes (DEGs) in Cmo-3 and Cma-2, respectively. Plant hormone signal transduction, Phenylpropanoid biosynthesis and the MAPK signaling pathway were found to be the significant KEGG pathways. The expression of ARF (auxin response factor), B-ARR (type-B response regulator) and PYR (pyrabactin resistance)/PYL (PYR-LIKE) genes was downregulated by NaCl treatment. In contrast, the expression of SnRK2 (sucrose non-fermenting-1-related protein kinase 2) and AHP (histidine-containing phosphotransmitter) genes was downregulated in Cmo-3 and upregulated in Cma-2. These findings will help us better understand the mechanisms of salt tolerance in pumpkins and potentially provide insight into enhancing salt tolerance in crop plants.