Neural Computing and Applications最新文献

To fulfill the explosion of multi-modal data, multi-modal sentiment analysis (MSA) emerged and attracted widespread attention. Unfortunately, conventional multi-modal research relies on large-scale datasets. On the one hand, collecting and annotating large-scale datasets is challenging and resource-intensive. On the other hand, the training on large-scale datasets also increases the research cost. However, the few-shot MSA (FMSA), which is proposed recently, requires only few samples for training. Therefore, in comparison, it is more practical and realistic. There have been approaches to investigating the prompt-based method in the field of FMSA, but they have not sufficiently considered or leveraged the information specificity of visual modality. Thus, we propose a vision-enhanced prompt-based model based on graph structure to better utilize vision information for fusion and collaboration in encoding and optimizing prompt representations. Specifically, we first design an aggregation-based multi-modal attention module. Then, based on this module and the biaffine attention, we construct a syntax–semantic dual-channel graph convolutional network to optimize the encoding of learnable prompts by understanding the vision-enhanced information in semantic and syntactic knowledge. Finally, we propose a collaboration-based optimization module based on the collaborative attention mechanism, which employs visual information to collaboratively optimize prompt representations. Extensive experiments conducted on both coarse-grained and fine-grained MSA datasets have demonstrated that our model significantly outperforms the baseline models.

The proposed paper introduces a new optimization algorithm inspired by nature called the polar fox optimization algorithm (PFA). This algorithm addresses the herd life of polar foxes and especially their hunting method. The polar fox jumping strategy for hunting, which is performed through high hearing power, is mathematically formulated and implemented to perform optimization processes in a wide range of search spaces. The performance of the polar fox algorithm is tested with 14 classic benchmark functions. To provide a comprehensive comparison, all 14 test functions are expanded, shifted, rotated and combined for this test. For further testing, the recent CEC 2021 test’s complex functions are studied in the unimodal, basic, hybrid and composition modes. Finally, the rate of convergence and computational time of PFA are also evaluated by several changes with other algorithms. Comparisons show that PFA has numerous benefits over other well-known meta-heuristic algorithms and determines the solutions with fewer control parameters. So it offers competitive and promising results. In addition, this research tests PFA performance with 6 different challenging engineering problems. Compared to the well-known meta-artist methods, the superiority of the PFA is observed from the experimental results of the proposed algorithm in real-world problem-solving. The source codes of the PFA are publicly available at https://github.com/ATR616/PFA.

This article is concerned with sampled-data synchronization problem of heterogeneous delays inertial neural networks (INNs) with generally uncertain semi-Markovian (GUSM) jumping. Different from traditional Markovian inertial neural networks (MINNs), the INNs with GUSM are investigated in this paper by fully considering the sojourn time and the lacking transition rates, which is more general and applicable for practical system. The new extended two-sided looped-functional (ETSLF) approach is adopted in this paper, and some improved less conservative criteria are derived to achieve the synchronization of the drive and response INNs. The controller gain matrices are acquired based on synchronization criteria. Finally, the viability of the method is presented through three examples.

In the exploration of recently developed Bayesian Generalized Nonlinear Models (BGNLM), this paper proposes a pragmatic scalable approximation for computing posterior distributions. Traditional Markov chain Monte Carlo within the populations of the Genetically Modified Mode Jumping Markov Chain Monte Carlo (GMJMCMC) algorithm is an NP-hard search problem. To linearize them, we suggest using instead variational Bayes, employing either mean-field approximation or normalizing flows for simplicity and scalability. This results in an evolutionary variational Bayes algorithm as a more scalable alternative to GMJMCMC. Through practical applications including inference on Bayesian linear models, Bayesian fractional polynomials, and full BGNLM, we demonstrate the effectiveness of our method, delivering accurate predictions, transparency and interpretations, and accessible measures of uncertainty, while improving the scalability of BGNLM inference through on the one hand using a novel variational Bayes method, but, on the other hand, enabling the use of GPUs for computations.

This paper explores the application of Visual Question-Answering (VQA) technology, which combines computer vision and natural language processing (NLP), in the medical domain, specifically for analyzing radiology scans. VQA can facilitate medical decision-making and improve patient outcomes by accurately interpreting medical imaging, which requires specialized expertise and time. The paper proposes developing an advanced VQA system for medical datasets using the Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation (BLIP) architecture from Salesforce, leveraging deep learning and transfer learning techniques to handle the unique challenges of medical/radiology images. The paper discusses the underlying concepts, methodologies, and results of applying the BLIP architecture and fine-tuning approaches for VQA in the medical domain, highlighting their effectiveness in addressing the complexities of VQA tasks for radiology scans. Inspired by the BLIP architecture from Salesforce, we propose a novel multi-modal fusion approach for medical VQA and evaluating its promising potential.

With the continuous emergence and rapid development of high and new technologies such as big data, cloud computing, artificial intelligence, mobile Internet, and the Internet of Things, the platform economy has developed rapidly and has become the current mainstream business model. This paper first analyzes the external driving factors that promote the rapid development of platform-based business models, then combines the existing research results of scholars to analyze the components of platform-based business models and uses deep learning methods. The research carried out model construction, drew causal relationship diagrams and flow diagrams, selected typical and representative platform-based enterprises for research, collected relevant data, and verified that the model's effectiveness reached 98%. On this basis, the model was compounded. Simulation and sensitivity analysis explores the critical factor driving platform-type enterprises to carry out business model innovation: the service quality coefficient of platform-type enterprises.

Accurate and reliable wind power forecasting is of great importance for stable grid operation and advanced dispatch planning. Due to the complex, non-stationary, and highly volatile nature of wind power data, Transformer-based methods find it difficult to capture long-term trend features and incur high computational costs. To address these challenging problems, we propose a frequency domain decomposition Transformer architecture with signal enhanced attention mechanism (SEAformer). Firstly, we devise a frequency domain-based trend decomposition structure that enables the Transformer to extract more effective long-term trend features, thereby further improving the long-term prediction accuracy of the model. Secondly, in response to the large fluctuations and instability of wind power data, we design an internal signal enhanced substructure combined with an attention mechanism in the Transformer, which filters out high-frequency noise signals and reduces the computational cost of the Transformer. We conduct extensive experiments on the benchmark dataset, the experimental analysis demonstrates that SEAformer outperforms the baseline methods (Transformer-based, MLP-based, and Traditional methods) in both multivariate and univariate prediction tasks, exhibiting the best prediction performance.

Projections of commodity prices have long been a significant source of dependence for investors and the government. This study investigates the challenging topic of forecasting the daily regional steel price index in the northeast Chinese market from January 1, 2010, to April 15, 2021. The projection of this significant commodity price indication has not received enough attention in the literature. The forecasting model that is used is Gaussian process regressions, which are trained using a mix of cross-validation and Bayesian optimizations. The models that were built precisely predicted the price indices between January 8, 2019, and April 15, 2021, with an out-of-sample relative root mean square error of 0.5432%. Investors and government officials can use the established models to study pricing and make judgments. Forecasting results can help create comparable commodity price indices when reference data on the price trends suggested by these models are used.

Ternary nanofluids have been an interesting field for academics and researchers in the modern technological era because of their advanced thermophysical properties and the desire to increase heat transfer rates. Furthermore, the innovative, sophisticated artificial neural network strategy with the Levenberg–Marquardt backpropagation technique (LMBPT) is proposed for research on heat and mass transport over non-Newtonian ternary Casson fluid on a radially extending surface with magnetic field and convective boundary conditions. The main objective of the current research is to conduct a comparative study of numerical solutions of the ternary nanofluid model of heat/mass transport utilizing the artificial neural network (ANN) together with the (LMBPT). To accurately represent complex patterns, neural networks modify their parameters flexibly, resulting in more accurate predictions and greater generalization with numerical outcomes. The model equations were reduced from partial to ODEs through applying appropriate similarity variables. The shooting technique and the byp-4c algorithm were then used to analyze the numerical data. The current study reveals that a rise in the Casson parameter diminishes the fluid velocity but an opposite nature is seen in thermal distribution for rising behavior of heat source/sink and Biot number, and the concentration profile tends to deteriorate when the mass transfer is elevated. Furthermore, the resulting values of the significant engineering coefficients are numerically analyzed and tabulated.

In this paper, we investigate the method of solving partial differential equations (PDEs) using artificial neural network (ANN) structures, which have been actively applied in artificial intelligence models. The ANN model for solving PDEs offers the advantage of providing explicit and continuous solutions. However, the ANN model for solving PDEs cannot construct a conventionally solvable linear system with known matrix solvers; thus, computational speed could be a significant concern. We study the implementation of the multigrid method, developing a general concept for a coarse-grid correction method to be integrated into the ANN-PDE architecture, with the goal of enhancing computational efficiency. By developing a reduced form of the multigrid method for ANN, we demonstrate that it can be interpreted as an equivalent representation of the classic multigrid expansion. We validated the applicability of the proposed method through rigorous experiments, which included analyzing loss decay and the number of iterations along with improvements in terms of accuracy, speed, and complexity. We accomplished this by employing the gradient descent method and the Broyden–Fletcher–Goldfarb–Shanno (BFGS) method to update the gradients while solving the given ANN systems of PDEs.