Large Number Of Data Samples Research Articles

Implementation and Efficient Analysis of Preprocessing Techniques in Deep Learning for Image Classification

Deep learning models have recently been preferred to perform certain image-processing tasks. Recently, with the increasing radiation, heat, and poor lighting conditions, the raw image samples may contain noisy and ambiguous information. To process these images, the deep learning model requires a large number of data samples to learn the missing information from other clear data samples. This necessitates training the neural network with a huge dataset. The researchers are now attempting to filter and improve such noisy images via preprocessing in order to provide valid and accurate feature information to the neural network layers. However, certain research studies claim that some useful information may be lost when the image is not preprocessed with an appropriate filter or enhancement technique. The MSA (meta-synthesis and analysis) approach is utilized in this work to present the impact of the image processing applications done with and without preprocessing steps. Also, this work summarizes the existing deep learning-based image processing models utilizing or not preprocessing steps in their implementation. This work has also found that 85% of the existing techniques involve a preprocessing step while developing a deep learning model. However, a maximum accuracy of 96.89% is observed on Sine-Net when it is implemented without a preprocessing and the same model gave 96.85% when implemented with preprocessing. This research provides various research insights on the requirement and non-requirement of preprocessing steps in a deep learning-based implementation.

Study on Rapid Simulation of the Pre-Cooling Process of a Large LNG Storage Tank with the Consideration of Digital Twin Requirements

The pre-cooling of a large LNG storage tank involves complex phenomena such as heat transfer, low-temperature flow, gas displacement, and vaporization. The whole pre-cooling process could take up to 50 h. For large-scale, full-capacity storage tanks, it is particularly important to accurately control the pre-cooling temperature. Digital twin technology can characterize and predict the full life cycle parameters from the beginning of pre-cooling development to the end and even the appearance of damage in real time. The construction of a digital twin platform requires a large number of data samples in order to predict the operating state of the device. Therefore, a simulation method with high computational efficiency for the pre-cooling process of LNG tanks is of great importance. In this paper, the mixture model and discrete phase model (DPM) are applied to simulate the pre-cooling process of a large LNG full-capacity tank. Following Euler–Lagrange, the DPM greatly simplifies the solution process. Compared with the experimental results, the maximum error of the DPM simulation results is less than 11%. Such a highly efficient simulation method for the large LNG full-capacity storage tank can make it possible to build the digital twin platform that needs hundreds of data model samples.

Study on the Influence of Home-school Cooperation on the Mental Health of College Students in China and its Mechanism

The purpose of this study is to explore the influence of home-school cooperation on the mental health of college students in China and its mechanism. By collecting a large number of sample data and using descriptive statistics, factor analysis and regression analysis, this paper makes an in-depth analysis of the relationship between home-school cooperation and college students' mental health. The results show that home-school cooperation has a significant positive impact on college students' mental health, among which the frequency of home-school communication, the diversity of home-school cooperation forms and the depth of home-school cooperation have a positive impact on college students' mental health. Further analysis shows that home-school cooperation can not only understand and solve students' psychological problems in time, but also promote students' psychological healthy growth by providing rich growth environment and personalized psychological support. Based on the above findings, this study puts forward some policy suggestions, such as strengthening home-school communication, enriching the forms of home-school cooperation, deepening the content of home-school cooperation and improving the policy support system, in order to provide more powerful protection for the mental health of college students in China. These research results not only help to enrich and develop the theoretical system in the field of home-school cooperation and college students' mental health, but also provide an important reference for educational practice and policy formulation.

Energy induced carbon nanotube rapidly assemble variable interweaving networks at the carbon fiber interface

Regulating carbon fiber with carbon nanotubes (CNTs) is important to improve the interface mechanical properties of composite materials. Therefore, a “thermotropic flash assembly (TFA)” process for the synthesis of CNTs in liquid phase is proposed. This work achieves the continuous and controllable formation of a micro-nano interweaving network of CNTs, with predictions made from a thermal perspective through enhancements to the process. Firstly, 12 working conditions and a large number of sample data are established by microcontrolling carbon fiber energy, and the growth thickness of CNTs is predicted by self-learning algorithm. Secondly, the “liquid-solid-solid” thermotropic growth of CNTs is proposed for the first time, which is different from the existing “gas-liquid-solid” and “gas-solid-solid” growth mechanisms of CNTs. Theoretical deductions and calculations are made for the interfacial temperature, exothermic, and endothermic energies during the growth process, which leads to the determination of a thermal growth space radius of 2.4 μm at the interface. Finally, it is found that the shear strength is increased by at least 27.7 %, although some tensile strength is sacrificed. This work not only points out the direction for low-cost, large-scale and customized control of interface properties, but also lays a theoretical foundation for exploring the growth mechanism of CNTs in an open environment at low temperature.

Construction of an innovative entrepreneurship project learning platform introducing a group recommendation algorithm for college students

This work aims to construct an innovative entrepreneurship project learning platform based on a group recommendation algorithm, which assists college students in better identifying and utilizing appropriate team resources during the entrepreneurial process. In order to achieve this goal, the genetic algorithm and probability models are introduced in the research process to improve the accuracy and personalization of group recommendations. First, the genetic algorithm is designed based on the initial population of individuals consisting of college students' diverse innovative entrepreneurship project information and personal characteristics. Individuals are iteratively optimized to find the optimal team combinations through crossover, mutation, and selection. Simultaneously, a probability model is employed to establish the group recommendation algorithm, and a large number of data samples and features are utilized for model training and prediction. Finally, the effectiveness of the proposed Genetic Algorithm-Probability Model-Group Recommendation (GAPMGR) is validated through experiments and comparative analysis. The research findings indicate that the GAPMGR method proposed has significantly improved team combination recommendations for college students' innovative entrepreneurship project learning platforms. Compared to traditional methods, the GAPMGR algorithm's group recommendation performance has improved by 4.92%, enabling more accurate matching of suitable team resources. By introducing the genetic algorithm and probability models, the innovative entrepreneurship project learning platform based on group recommendation can effectively enhance students' entrepreneurial abilities and team collaboration effects. This platform can provide personalized team resource recommendations for students, promote interdisciplinary cooperation, and facilitate the successful implementation of innovative entrepreneurship projects. In the fields of education and entrepreneurship, this work provides a valuable entrepreneurial learning tool for college students and serves as a reference for educational institutions and entrepreneurship mentors. In the field of technology, this work explores the application of the genetic algorithm and probability models in group recommendation, offering new ideas and methods for further research and application.

Remote field-programmable gate array laboratory for signal acquisition and design verification

A remote laboratory utilizing field-programmable gate array (FPGA) technologies enhances students’ learning experience anywhere and anytime in embedded system design. Existing remote laboratories prioritize hardware access and visual feedback for observing board behavior after programming, neglecting comprehensive debugging tools to resolve errors that require internal signal acquisition. This paper proposes a novel remote embedded-system design approach targeting FPGA technologies that are fully interactive via a web-based platform. Our solution provides FPGA board access and debugging capabilities beyond the visual feedback provided by existing remote laboratories. We implemented a lab module that allows users to seamlessly incorporate into their FPGA design. The module minimizes hardware resource utilization while enabling the acquisition of a large number of data samples from the signal during the experiments by adaptively compressing the signal prior to data transmission. The results demonstrate an average compression ratio of 2.90 across three benchmark signals, indicating efficient signal acquisition and effective debugging and analysis. This method allows users to acquire more data samples than conventional methods. The proposed lab allows students to remotely test and debug their designs, bridging the gap between theory and practice in embedded system design.

Hypergraph construction using Multi-Sensor for helicopter Tail-Drive system fault diagnosis

Recently, neural network research in the field of mechanical fault diagnosis often requires a large number of data samples, while graph neural networks constructed using a single sensor under limited samples of the graph topological relationship is difficult to cover the coupling information, and the problem of over-smoothing of the graph can lead to failure of the fault diagnosis. As a result, Multi-Channel Hypergraph Convolutional Network (McHGCN) for limited-sample fault diagnosis of helicopter tail drive system is proposed. Features are extracted from multi-sensor data and hyper edges are constructed by k-nearest neighbor method thus exploiting the similarity. Nodal feature representation of multi-sensor data is done with 1D convolutional network, which prevents over-smoothing while preserving the learning depth. Finally, fault diagnosis is accomplished by semi-supervised learning of the hypergraph convolutional layer. Limited-sample fault diagnosis experiments are conducted on a helicopter tail drive system to verify the effectiveness of the method.

Load Calculation Method for Deep-Buried Layered Soft Rock Tunnel Based on Back-Analysis of Structural Deformation

After the excavation and unloading of deep-buried soft rock tunnels, support structures often experience deformation-related disasters such as concrete cracking, steel frame bending and twisting, and primary support instability under different forms of load. Accurately calculating the load borne by the primary support structure is the key to ensuring design rationality and construction safety. Especially in layered soft surrounding rock formations, the magnitude and distribution of the loads are different from those of conventional rock and soil masses, resulting in limited applicability of existing load calculation methods to similar formations. Therefore, based on the measured deformation of the tunnel structure, while considering the different geometric forms of the primary support structure during partial excavation, this paper proposes a deformation-structure (D-S) load calculation method. By comparing the calculation results of this method and a large number of sample data for typical deep-buried layered soft rock tunnels, the reliability of the D-S load calculation method is verified. In addition, the variation law of the loads during the tunnel construction period is enunciated, and the magnitude and distribution of the loads acting on the primary support are clarified. The D-S load calculation method provides a theoretical basis for load calculation in deep-buried layered soft rock tunnels.

Optimization of the selection of suitable harvesting periods for medicinal plants: taking Dendrobium officinale as an example

BackgroundDendrobium officinale is a medicinal plant with high commercial value. The Dendrobium officinale market in Yunnan is affected by the standardization of medicinal material quality control and the increase in market demand, mainly due to the inappropriate harvest time, which puts it under increasing resource pressure. In this study, considering the high polysaccharide content of Dendrobium leaves and its contribution to today’s medical industry, (Fourier Transform Infrared Spectrometer) FTIR combined with chemometrics was used to combine the yields of both stem and leaf parts of Dendrobium officinale to identify the different harvesting periods and to predict the dry matter content for the selection of the optimal harvesting period.ResultsThe Three-dimensional correlation spectroscopy (3DCOS) images of Dendrobium stems to build a (Split-Attention Networks) ResNet model can identify different harvesting periods 100%, which is 90% faster than (Support Vector Machine) SVM, and provides a scientific basis for modeling a large number of samples. The (Partial Least Squares Regression) PLSR model based on MSC preprocessing can predict the dry matter content of Dendrobium stems with Factor = 7, RMSE = 0.47, R2 = 0.99, RPD = 8.79; the PLSR model based on SG preprocessing can predict the dry matter content of Dendrobium leaves with Factor = 9, RMSE = 0.2, R2 = 0.99, RPD = 9.55.ConclusionsThese results show that the ResNet model possesses a fast and accurate recognition ability, and at the same time can provide a scientific basis for the processing of a large number of sample data; the PLSR model with MSC and SG preprocessing can predict the dry matter content of Dendrobium stems and leaves, respectively; The suitable harvesting period for D. officinale is from November to April of the following year, with the best harvesting period being December. During this period, it is necessary to ensure sufficient water supply between 7:00 and 10:00 every day and to provide a certain degree of light blocking between 14:00 and 17:00.

Data privacy protection: A novel federated transfer learning scheme for bearing fault diagnosis

Research on the health diagnosis of mechanical equipment has developed unprecedentedly in recent years, and a large number of diagnostic solutions have considerably improved the stability of mechanical equipment in industrial production. However, such satisfactory diagnostic performance relies on a large number of data samples, which are frequently difficult to obtain in real industrial scenarios. The traditional strategy of data sharing is no longer advisable due to the potential conflict of interest among users. A federated transfer learning scheme is proposed to alleviate the data island problem in industrial production while protecting data privacy. This solution adopts a distributed structure, which includes local model training and global model update. A differential training scheme is proposed to enhance the domain adaptability of the local model. The central server evaluates the contribution ability of each local model to the target task. It also weights and aggregates each client model on the basis of parameter importance ranking in the form of model fusion. The target task of the experiment is performed on two sets of bearing datasets. By comparing with other diagnostic methods, a conclusion can be drawn that the proposed scheme provides a promising federated learning method while protecting client data privacy.

Private Equity Investment and Corporate Technological Innovation: An Empirical Examination Based on Chinese Mainboard-listed High-Tech

Technological innovation is very important to the growth and success of an enterprise, which greatly affects its market competitiveness. There are many known factors that affect the technological innovation of enterprises, and private equity investment is one of the important but often neglected factors. As a country with relatively developed finance, China's laws and systems related to private equity investment are relatively sound. Based on a large number of sample data of high-tech enterprises listed on the main board of China in 2012-2022, this paper studies the influence of private equity investment on technological innovation of enterprises. This paper finds that private equity investment has an obvious positive effect on enterprise technological innovation, and the effect of private equity investment on innovation shows certain differences due to different types of enterprises, which is embodied in the fact that state-owned enterprises or smaller enterprises are more sensitive to private equity investment. The research has both theoretical and practical significance, and can provide some help and suggestions for enterprises and the government in making decisions related to private placement.

Electric Vehicle and Photovoltaic Power Scenario Generation under Extreme High-Temperature Weather

In recent years, with the intensification of global warming, extreme weather has become more frequent, intensifying the uncertainty of new energy output and load power, and seriously affecting the safe operation of power systems. Scene generation is an effective method to solve the uncertainty problem of stochastic planning of integrated systems of new energy generation. Therefore, this paper proposes a scenario generation and scenario reduction model of photovoltaic (PV) output and electric vehicle (EV) load power under extreme weather based on the copula function. Firstly, the non-parametric kernel density estimation method is used to fit a large number of sample data. The kernel density estimation expressions of PV and EV powers under extreme weather conditions are obtained and the corresponding goodness of fit tests are carried out. Then, a variety of joint distribution models based on the copula function are established to judge the goodness of fit of each model, and the optimal copula function is selected as the joint probability distribution function by combining the Kendall and Spearman correlation coefficients of each model. Finally, the optimal copula joint probability distribution is used to generate PV and EV power scenarios. The data of extremely hot weather in a certain province were selected for an example analysis. The results show that the output scenario obtained conforms to the correlation under this extreme weather, and has higher accuracy in reflecting the actual PV output and load power in this province under this extreme weather, which can provide a reference for reliability analyses of power systems and power grid planning.

Learning Disentangled Representation for One-Shot Progressive Face Swapping.

Although face swapping has attracted much attention in recent years, it remains a challenging problem. Existing methods leverage a large number of data samples to explore the intrinsic properties of face swapping without considering the semantic information of face images. Moreover, the representation of the identity information tends to be fixed, leading to suboptimal face swapping. In this paper, we present a simple yet efficient method named FaceSwapper, for one-shot face swapping based on Generative Adversarial Networks. Our method consists of a disentangled representation module and a semantic-guided fusion module. The disentangled representation module comprises an attribute encoder and an identity encoder, which aims to achieve the disentanglement of the identity and attribute information. The identity encoder is more flexible, and the attribute encoder contains more attribute details than its competitors. Benefiting from the disentangled representation, FaceSwapper can swap face images progressively. In addition, semantic information is introduced into the semantic-guided fusion module to control the swapped region and model the pose and expression more accurately. Experimental results show that our method achieves state-of-the-art results on benchmark datasets with fewer training samples. Our code is publicly available at https://github.com/liqi-casia/FaceSwapper.

Research on a High-Performance Rock Image Classification Method

Efficient and convenient rock image classification methods are important for geological research. They help in identifying and categorizing rocks based on their physical and chemical properties, which can provide insights into their geological history, origin, and potential uses in various applications. The classification and identification of rocks often rely on experienced and knowledgeable professionals and are less efficient. Fine-grained rock image classification is a challenging task because of the inherent subtle differences between highly confusing categories, which require a large number of data samples and computational resources, resulting in low recognition accuracy, and are difficult to apply in mobile scenarios, requiring the design of a high-performance image processing classification architecture. In this paper we design a knowledge distillation and high-accuracy feature localization comparison network (FPCN)-based learning architecture for generating small high-performance rock image classification models. Specifically, for a pair of images, we interact with the feature vectors generated from the localized feature maps to capture common and unique features, let the network focus on more complementary information according to the different scales of the objects, and then the important features of the images learned in this way are made available for the micro-model to learn the critical information for discrimination via model distillation. The proposed method improves the accuracy of the micro-model by 3%.

Feature extraction algorithm of anti-jamming cyclic frequency of electronic communication signal

Abstract Anti-jamming cyclic frequency feature extraction is an important link in identifying communication interference signals, which is of great significance for eliminating electronic communication interference factors and improving the security of electronic communication environment. However, when the traditional feature extraction technology faces a large number of data samples, the processing capacity is low, and it cannot solve the multi-classification problems. For this type of problem, a method of electronic communication signal anti-jamming cyclic frequency feature extraction based on particle swarm optimization-support vector machines (PSO-SVM) algorithm is proposed. First, the SVM signal feature extraction model is proposed, and then the particle swarm optimization (PSO) algorithm is used. Optimize the kernel function parameter settings of SVM to raise the classifying quality of the SVM model. Finally, the function of the PSO-SVM signal feature extraction model is tested. The results verify that the PSO-SVM model begins to converge after 60 iterations, and the loss value remains at about 0.2, which is 0.2 lower than that of the SVM technique. The exactitude of signal feature extraction is 90.4%, and the recognition effect of binary phase shift keying signal is the best. The complete rate of signal feature extraction is 85%. This shows that the PSO-SVM model enhances the sensitivity of the anti-jamming cyclic frequency feature, improves the accuracy of the anti-jamming cyclic frequency feature recognition, reduces the running process, reduces the time cost, and greatly increases the performance of the SVM method. The good model performance also improves the application value of the method in the field of electronic communication.

Deterministic bridge regression for compressive classification

Pattern classification using a compact representation is a crucial component of machine intelligence. Specifically, it is essential to learn a model with well-regulated parameters to achieve good generalization. Bridge regression provides a mechanism for regulating parameters through a penalized ℓp-norm. However, due to the nonlinear nature of the formulation, an iterative numerical search is typically used to solve the optimization problem. In this work, we propose an analytic solution for bridge regression based on solving a penalized error formulation using an approximated ℓp-norm. The solution is presented in primal form for over-determined systems and in dual form for under-determined systems. The primal form is suitable for low-dimensional problems with a large number of data samples, while the dual form is suitable for high-dimensional problems with a small number of data samples. We also extend the solution to problems with multiple classification outputs. Numerical studies using simulated and real-world data demonstrate the effectiveness of our proposed solution.

Maximum power point tracking control of wind turbines based on equivalent sinusoidal wind

The optimal parameter model method for maximum power point tracking (MPPT) control of wind turbines has good wind energy capture performance. However, there is a precondition that the quantitative relationship between optimal control parameters and turbulence characteristics is accurately described. The resolution of the existing turbulence frequency characterization indicators are insufficient. Therefore, a large number of sample data need to be obtained to clearly show the statistical relationship between them and the optimal control parameters, which is extremely time-consuming, and has a negative impact on the optimization results of control parameters. This paper equates the turbulent wind to the sinusoidal wind with the objective of the results of both acting on wind turbines are the same, so as to realize the extraction and index description of the dominant turbulent frequency components strongly related to the optimal control parameters. On this basis, a MPPT control method based on equivalent sinusoidal wind is proposed. This method greatly reduces the required sample size, and makes the setting of control parameters closer to the optimal value, which can further improve the efficiency of wind energy capture. Simulation analysis and experimental results based on wind turbine simulator verify the effectiveness of the proposed method.

Deep learning based on model migration for COVID-19 identification

Since the end of 2019, the virus has gradually spread and eventually spread globally. In this context, it is important to control the spread of COVID-19 quickly. This project attempts to use artificial intelligence to identify CT images of the lungs of COVID-19 patients and facilitate rapid screening of COVID-19 patients. The main focus of this study is to use artificial intelligence based on model transfer deep learning to identify whether patients are infected with novel coronavirus through patient lung images. The difficulty of this task is that the number of lung images of COVID-19 patients is very limited, which makes it very difficult to train traditional neural networks. Traditional computer vision deep learning to extract image features requires a large number of sample data for model training. If the number of images in the data set is too small, the model will overfit and fail to achieve relatively accurate COVID-19 identification effect. To solve the above problems, this paper studied novel coronavirus identification of patients' lung CT images by deep learning method based on model transfer. We build models based on similar types of problems, store those models and then fine-tune them. Eventually, a model was trained to recognize images of the lungs of COVID-19 patients. The method was tested on publicly available COVID-19 datasets, and the results showed that the identification accuracy of the method was about 70%.

NFSP-PLT: Solving Games with a Weighted NFSP-PER-Based Method

Nash equilibrium strategy is a typical goal when solving two-player imperfect-information games (IIGs). Neural fictitious self-play (NFSP) is a popular method to find the Nash equilibrium in IIGs, which is the first end-to-end method used to compute the Nash equilibrium strategy. However, the training of NFSP requires a large number of sample data and the interactive cost of obtaining such data is often very high. Realizing the efficient training of network under limited samples is an urgent problem. In this paper, we first proposed a new NFSP-based method, NFSP with prioritized experience replay (NFSP-PER), to improve the sample training efficiency. Then, a weighted NFSP-PER with learning time (NFSP-PLT) was proposed to control the utilization degree of priority-weighted samples. Furthermore, based on the NFSP-PLT, an adaptive upper-confidence-bound applied to tree (UCT) is used to solve the optimal response strategy, which makes the solving strategy more accurate. Extensive experimental results show that the proposed NFSP-PLT effectively improves the sample learning efficiency compared with the existing works.

Improving Recognition Accuracy of Pesticides in Groundwater by Applying TrAdaBoost Transfer Learning Method.

Accurate and rapid prediction of pesticides in groundwater is important to protect human health. Thus, an electronic nose was used to recognize pesticides in groundwater. However, the e-nose response signals for pesticides are different in groundwater samples from various regions, so a prediction model built on one region's samples might be ineffective when tested in another. Moreover, the establishment of a new prediction model requires a large number of sample data, which will cost too much resources and time. To resolve this issue, this study introduced the TrAdaBoost transfer learning method to recognize the pesticide in groundwater using the e-nose. The main work was divided into two steps: (1) qualitatively checking the pesticide type and (2) semi-quantitatively predicting the pesticide concentration. The support vector machine integrated with the TrAdaBoost was adopted to complete these two steps, and the recognition rate can be 19.3% and 22.2% higher than that of methods without transfer learning. These results demonstrated the potential of the TrAdaBoost based on support vector machine approaches in recognizing the pesticide in groundwater when there were few samples in the target domain.