最新刊期
卷 22 ， 期 2 ， 2023
- 摘要：Aiming at the problems of scarcity of ethnic medicinal plants image data set， small sample size complex image background， which make image feature extraction difficult， the TibetanMP data set is constructed and an image recognition method with embedded Squeeze-and-Excitation mechanism ResNet combined with transfer learning was proposed. In this method， the pre-training model of ResNet34 on ImageNet is transferred to reduce the over-fitting phenomenon. Meanwhile， SE mechanism is introduced in the shallow layer of the network to focus the key features in the image. Finally， the model is fine-tuned. In order to evaluate the performance of the proposed method， on the TibetanMP， Oxford 102 flowers and CIFAR-10 datasets， the model achieved recognition accuracy of 96.33%， 98.81% and 91.92%， respectively. Compared with other mainstream CNN image recognition models， this model has higher recognition accuracy. Experiments show that this method can effectively improve the image recognition performance of ethnic medicinal plants， and has certain engineering practicability.关键词：image recognition;ethnic medicinal plants images;Resnet;squeeze and excitation;transfer learning
- 242
- |
- 80
- |
- 0
发布时间：2023-03-01 - 摘要：Color image edges play a vital role in computer vision， but the existing detection algorithms have poor resistance to mixed noise， poor continuity of detected edges， and missed edges. Through the research of color image edge detection algorithm， an improved color image edge detection algorithm is proposed. The algorithm first uses the proposed improved wavelet transform filtering method to filter the mixed noise of images while maintaining the edges. Then the color image after filtering is converted from RGB to the CLElab color space， and the multidirectional color difference formula is proposed to calculate color gradient amplitude and phase angle for improving the completeness of the detected edges. Finally， the color image edge detection is accomplished by combining the above processing methods with the non-maximum suppression and double-threshold edge connection of the Canny method. The experimental results show that the proposed algorithm has high recall， accuracy and overall accuracy， reaching 85.9%， 96.7% and 96.3%，and compared with the existing color difference edge detection algorithm， brightness and color feature fused edge detection algorithm and improved colony based edge detection algorithm， it is improved by 3.8% to 14.8%， 1.6% to 8%， and 1.3% to 3.7%， respectively. The improved Canny-based color image edge detection algorithm can effectively remove mixed noise and detect color edges with high continuity and integrity， thus effectively improving the edge detection rate.关键词：color image edge detection;Canny;wavelet transform;edge-preserving filtering;color difference calculation
- 139
- |
- 47
- |
- 0
发布时间：2023-03-01 - 摘要：Organ lesions have a high mortality rate and seriously threaten the safety of human life. The internal organs of human body are diverse in form and complex in anatomical structure，accurate segmentation of the organ assists the doctor in making the diagnosis. High precision segmentation model is required for medical image. However， most segmentation models are directly transferred from the general image segmentation model. These models often ignore the importance of shallow feature information and boundaries. In order to solve this problem， attention mechanism and point sampling technique are proposed to obtain high quality segmentation boundary. The model was evaluated on CHAOS， a commonly used liver medical image dataset， and the average Dice was 0.946 7， the average IoU was 0.962 3， and the average F1 Score was 0.9351. It is proved that this model can learn both the detail features and the global structure features of the image， and can perform better segmentation of the liver image.关键词：medical image segmentation;U-Net;attention gate;point sampling technique
- 66
- |
- 25
- |
- 0
发布时间：2023-03-01 - 摘要：In order to solve the problem that the attention mechanism model of convolution neural networks （CNN） can only focus on local features and small receptive field when reconstructing infrared images with a wide wavelength range， propose a new method with lightweight ViT and CNN suitable for reconstructing infrared images with a wide range. The model used an improved lightweight residual block combined with a lightweight ViT block to build a global self-attention mechanism model， learned long-distance attention dependencies between different feature map regions to assist reconstruction and constrain the solution space. It used Huber loss function to make the model converge stably. It mined the deep transformation relationship between high and low resolution image pairs by iterative up and down sampling. Near-infrared images and far-infrared images datasets were used in the experiment， the model with 1 031K parameters surpassed the lightweight model SRResNet with 1 518K parameters and CARN with 1 592K parameters in the comparison of peak signal-to-noise ratio and structural similarity， close to the heavyweight model EDSR with a parameter amount of 4 543K，which shows that the model can effectively reconstruct infrared images with a wide wavelength range.关键词：infrared image;lightweight;ViT;super-resolution;self-attention
- 77
- |
- 24
- |
- 0
发布时间：2023-03-01 - 摘要：As a hot research problem in the field of image processing， image matting is widely used in target recognition， virtual reality and foreground extraction. However， it is always difficult to obtain high-quality pixel pairs when there are few high-quality pixels in the image. Therefore， this paper proposes an image matting algorithm based multi criteria sampling， it uses a multi criteria sampling strategy to obtain high-quality candidate subsets from the perspective of global similarity to local similarity， to solve the problem that high-quality pixel pairs are easy to lose when there are fewer high-quality pixel pairs. On this basis， a strategy for selecting the optimal pixel pair with multiple evaluations is designed，it combines multiple pixel pair evaluation functions to avoid the problem that the solution estimated by a single evaluation function is not the current optimal solution. In order to verify the superiority of the image matting algorithm based multi criteria sampling， the Alpha matting benchmark dataset is selected as the experimental data. The experimental results show that， compared with the current popular matting algorithm， the foreground alpha mattes estimated by the proposed algorithm is optimal.关键词：multi-criterion sampling;combination evaluation function;sampling matting algorithm;picture processing
- 44
- |
- 15
- |
- 0
发布时间：2023-03-01 - 摘要：Aiming at the problem that using two-dimensional image processing to measure the area of folds on the surface of tobacco leaves is not accurate， which affects the accuracy of tobacco leaf intelligent classification， a three-dimensional measurement method of tobacco leaf surface area based on binocular vision is proposed.First， use a binocular stereo vision system to collect tobacco leaf pictures and preprocess them， use the SGBM stereo matching algorithm to obtain the disparity map， and use morphological processing to post-process the disparity map；Then， use the principle of triangulation to obtain a three-dimensional point cloud， and import the point cloud data into Meshlab software to use the Ball Pivoting algorithm to perform three-dimensional reconstruction of the tobacco leaf surface；Finally， the surface area of the tobacco leaf is calculated by VTK statistical grid number.By comparing the surface area measurement experiments of 30 tobacco leaves in the upper， middle and lower parts， the results show that the proposed method improves the accuracy of tobacco leaf area measurement by about one-third.And it provides a corresponding basis for the analysis of the wrinkle characteristics in the subsequent intelligent grading process of tobacco leaves.关键词：binocular vision;stereo matching;3D point cloud;3D reconstruction;tobacco fold area
- 25
- |
- 4
- |
- 0
发布时间：2023-03-01 - 摘要：In order to solve the problem of complex Inception structure and redundant parameters in deep network in image classification， an improved Inception structure is proposed， which integrates the complicated 1×1 convolution operation in the traditional Inception structure， and introduces depthwise convolution. Increase the feature diversity of the Inception structure， reduce the amount of parameters， and combine the residual structure to prevent gradient explosion and gradient disappearance. At the same time， an attention mechanism is introduced to obtain key feature weight information and optimize resource allocation. On this basis， a lightweight network model based on standard convolution and asymmetric convolution is designed. In the experiment， the apple leaf disease dataset and the CIFAR-10 dataset were selected as the experimental objects. The results after comparing with the classic convolutional neural network VGG16， Inception-V3 and MobileNet show that the lightweight model of the improved Inception structure has fewer parameters， the advantages of short training time and better classification effect prove that the quality of the model proposed is better than that of VGG16， Inception-V3 and MobileNet. At the same time， by comparing the training results of the Inception structure before and after the improvement in the proposed network， it is also proved that the improved Inception structure is better than the traditional Inception structure.关键词：image classification;Inception;deep convolution;apple foliage disease;convolutional neural network
- 26
- |
- 8
- |
- 0
发布时间：2023-03-01 - 摘要：Aiming at the problems of incorrect entity boundary segmentation， overlapping entity relations， and the inability to transfer the relation types of different datasets， an entity relation extraction method based on Schema enhancement is proposed. First of all， the semantic information of characters and words is merged by word mixing and embedding to avoid the ambiguity problem caused by the error of boundary segmentation in Chinese word segmentation. Secondly， the problem of relationship overlap is solved by using pointer annotation. Finally， the Schema of each dataset is extracted and merged into the model as a priori features to solve the problem of entity redundancy and relationship type migration. Experiments were carried out on the three primary Chinese entity relation extraction datasets DuIE， FinRE， and SanWen. Compared with the previous model， they achieved F1 improvements of 10%， 18%， and 11%， respectively， and showed higher stability.关键词：named entity recognition;relation extraction;Schema enhancement;character-word embedding;pointer labeling
- 34
- |
- 9
- |
- 0
发布时间：2023-03-01 - 摘要：A deep learning model based on a variational autoencoder neural network learns and visualizes low-dimensional embedded representations of mass spectrometry images to reveal hidden organizational structures. Unsupervised analysis and peak learning were performed on mass spectrometry imaging data of mouse kidney tissue using a deep learning network framework. The msiPL approach learns and visualizes the underlying nonlinear spectral manifold， reveals biologically relevant clusters of mouse kidney tissue anatomy and tumor heterogeneity in a mouse gastric cancer model， and identifies potentially specific m/z peaks. This method can quickly and efficiently analyze mass spectrometry imaging data sets without peak picking.关键词：mass spectrometry imaging;unsupervised analysis;peak learning
- 22
- |
- 3
- |
- 0
发布时间：2023-03-01 - 摘要：Aiming at the problem that the number of partial discharge gray-scale images for power cables is small， and it is difficult to train deep residual network model based on large-scale data sets. A method of partial discharge pattern recognition for cables under few samples condition is proposed. The method uses the idea of combining expanded samples with simplified model. A data augmentation method based on deep convolutional generative adversarial network is proposed for the partial discharge grayscale image. In order to verify the effectiveness of the generated samples， an evaluation index of the generated samples based on the box dimension is proposed. The effects of residual modules and network depth on the classification performance of residual networks are compared. A simplified residual network model is proposed to match the small-scale partial discharge data set. The simplified residual network model is tested by experiments. The average iteration time of the network model is 7.3s， and the recognition accuracy rate reaches 98.5%. Compared with the method of directly using few samples to train the deep residual network， the proposed method has faster model training speed and higher recognition accuracy.关键词：generative adversarial network;residual network;partial discharge;pattern recognition;deep learning
- 21
- |
- 5
- |
- 0
发布时间：2023-03-01 - 摘要：In recent years， the simultaneous localization and mapping （SLAM） method dominated by visual sensors has received extensive attention and research. However， due to the real-time performance and accuracy of localization， a certain frequency of visual computing and map refresh needs to be guaranteed， and the demand for computing and storage is high， causing platforms with limited resources lose support as the project is updated iteratively. Therefore， propose a monocular vision SLAM method for ARM isomorphic processors. The method first constructs an up-to-scale initial map based on pure vision， and solves the visual scale factor by aligning the IMU measurement data with absolute scale. Then， a fast tracking strategy is proposed to improve the running speed by reducing the calculation amount of feature extraction and descriptors. Finally， the sliding window algorithm is used to limit the scale of the back-end optimization problem， and at the same time， the edge push and data integration of historical data are carried out in time， which effectively avoids the surge of local computing and storage under long-term operation. Experiments on the TUM visual inertial navigation datasets show that compared with the ORB-SLAM3 method， the average computing speed of this method on the Raspberry Pi is increased by 69.29%， and the absolute trajectory error is lower， supporting real-time push and offline map data， which is an effective method for the application of visual SLAM in the embedded system platform.关键词：SLAM;monocular vision;embedded devices;ARM;Raspberry Pi
- 22
- |
- 3
- |
- 0
发布时间：2023-03-01 - 摘要：At present， tomato picking mainly relies on manual labor， so it is urgent to realize the mechanization and intelligence of the tomato industry， and tomato detection is the most basic and most important step. In response to this， propose a tomato detection algorithm based on improved Mask RCNN. The algorithm selects ResNet50 and FPN as the backbone network， proposes a novel RoI extractor， and uses atrous convolution （Atrous） in the algorithm model. Through the Labelme self-made tomato data set， the improved algorithm will be trained and tested on the self-made data set. Compared with the Faster RCNN and Mask RCNN models， the improved model also increases the AP value by 5.5% and 4.7%， respectively， and the AR value that's an increase of 6.8% and 4.6%， respectively.The results show that it not only improves the recognition accuracy of tomatoes， but also better achieves instance segmentation.关键词：tomato detection;Mask RCNN;new RoI extractor;Atrous;instance segmentation
- 62
- |
- 12
- |
- 0
发布时间：2023-03-01 - 摘要：Facial expression recognition is one of the core of human-computer interaction research field. The existing facial expression recognition methods based on traditional manual features are difficult to be applied in complex and changeable application scenes. Based on this， an expression recognition method of DenseNet network with multi-scale attention mechanism is proposed. The network model simplifies the layers of DenseNet121 network， inserts multi-scale structure and channel attention module MECANet， which makes the facial expression features extracted by the network more discriminative and conducive to the expression classification of subsequent networks.The network model is trained by random gradient descent algorithm. High recognition rates are achieved on CK + and FER2013 data sets， reaching 96.2% and 85.5% respectively， which are 8.4% and 8.6% higher than DenseNet121 network.关键词：facial expression recognition;DenseNet;multiscale structure;MECANet
- 58
- |
- 8
- |
- 0
发布时间：2023-03-01 -
Extraction of Consistency Degradation Feature for Rolling Bearings Based on Cointegration Theory AI导读
摘要：Through the feature extraction of bearing whole-life data under different working conditions in multiple laboratories， it is analyzed that there is a certain cointegration relationship between the first difference absolute sum and the 0.1 quantile， and a consistence degradation feature extraction method for rolling bearings based on cointegration theory is proposed.This method can unify the whole life data of different rolling bearings， and obtain the consistence degradation feature under different working conditions of different bearings.The reason for the consistency of the proposed feature is analyzed by experiments on the whole-life data of several laboratories under different working conditions. Further， the similarity between feature sequences is calculated by dynamic time warping （DTW） method， the advantages of the proposed features in consistency are verified； The calculated results show that the consistency of the proposed cointegration feature is 2.024 times of the first difference absolute sum and 3.799 times of 0.1 quantile when measured by the mean of similarity degree.Compared with other common fusion algorithms， the results show that the proposed feature can effectively reduce the influence of factors such as working conditions， and has a good two-stage property.关键词：rolling bearing;cointegration theory;degradation feature;two-stage property;consistency- 15
- |
- 2
- |
- 0
发布时间：2023-03-01 - 摘要：The content retrieval process of content center network （CCN） is inefficient and redundant， In order to improve the performance of CCN， proposes a CCN community partition scheme （SI-LPA） based on node similarity and influence. Firstly， a community partition scheme based on feature vector centrality and improved label propagation algorithm is proposed to help content retrieval. Then， using the similarity and influence between nodes and neighbor nodes， a method to update the tag value is proposed； Finally， select the most competitive node on each divided community to deploy SDN controller to help better manage community functions and route between communities. Experiments show that the introduction of SDN controller and community partition can accelerate the speed of content retrieval and routing distribution， so as to improve the performance of CCN routing.关键词：CCN;community division;node similarity;node influence;SDN
- 21
- |
- 6
- |
- 0
发布时间：2023-03-01 - 摘要：Parameterization and re-parameterization of curves and surfaces is a basic problem in computer-aided geometric design. The approximate arc angle re-parameterization is studied for a class of rational curve parameter representation of angular velocity function with zero on unit interval. Based on the piecewise radical transformation and the piecewise Möbius transformation combined with the C1 arc angle new parameterization algorithm， a re-parameterization algorithm based on the C1 piecewise radical transformation was proposed. Experiments show that this algorithm can greatly improve the angular velocity uniformity represented by the rational parameters.关键词：rational parameter representation;approximate arc angle re-parameterization;C1 piecewise radical transformation
- 11
- |
- 2
- |
- 0
发布时间：2023-03-01 - 摘要：At present， the prevention and control of the COVID-19 epidemic in China has entered the stage of normalization management. However， due to factors such as overseas imports and cold chain transmission， the epidemic still has a sudden spread in some areas， resulting in an increase in the risk level of some urban areas， and the risk level of passengers who have been active in this area has also changed. On the premise of knowing the risk level of passengers， aiming at how to reduce the risk of passengers infected with virus during carpooling and achieve the maximum total social benefit of the vehicle， the constraint conditions between carpooling passengers are added， and an optimization model of carpooling scheduling is established considering the risk level of passengers. Based on the model， a small-scale numerical example is designed and solved by Lingo software， and the effectiveness of the model is analyzed and verified by numerical examples. The study found that the results of the example showed that the model could provide different car-sharing schemes according to the risk level of passengers. After carrying out different subsidy experiments on vehicles carrying high-risk passengers， it was found that with the increase of subsidy amount， the number of vehicles carrying high-risk passengers also increased significantly. This study has certain reference value for the optimization of carpooling scheduling in the epidemic situation.关键词：ride-sharing;passenger risk rating;COVID-19;scheduling optimization model;contingency management;smart transportation
- 23
- |
- 4
- |
- 0
发布时间：2023-03-01 - 摘要：By analyzing the various stages of cloud gaming’s activities and business latency， design a WebRTC cloud gaming platform system. Based on establishing the experience evaluation model， the congestion control algorithm to reduce the delay is studied deeply and implemented in code to send frame rate dynamically.The data shows that the algorithm can solve the problems such as business latency and lag caused by congestion， and reduce the end-to-end latency and lag rate of the platform to about 100ms and 4.13% respectively， which bring users cloud gaming services with low response delay and high streaming quality.关键词：cloud gaming;WebRTC;low-latency streaming;FPS;lag rate
- 57
- |
- 5
- |
- 0
发布时间：2023-03-01 - 摘要：In order to solve the complicated management problems in the process of raising pets， an intelligent pet management system based on GPS and WeChat applet is proposed. Based on the MINA framework， the system is designed using MySQL database， including home page， life record， appointment， pet file， navigation， electronic fence and other modules. It has the functions of recording pet habits， storing pet file information， appointment arrangement， locating pet locations and so on. Users mark a number of points in the electronic fence module to generate a polygon fence to set a safe range of pet activities to prevent pet loss. The practice results show that the system reduces the workload of pet keepers， improves the scientificity and safety of pet management， and provides new ideas for the field of pet management.关键词：pet management;WeChat applet;database;data interaction
- 85
- |
- 10
- |
- 0
发布时间：2023-03-01 - 摘要：Near-infrared spectroscopy （NIRS） and chemometrics were used for the identification of cyan nephrite from Luodian China， Qinghai China， Xinjiang China， and South Korea. The absorption peaks of 4 900~7 300 cm-1 bands were selected to calculate the peak area by multi-peak fitting method， and then the SPSS 25.0 statistical software was used to process the peak area data， and the identification model was established. The purpose is to explore the traceability method of different origins of cyan nephrite. The results show that the absorption of Luodian cyan nephrite at 5 100 and 6 831 cm-1 by NIRS is different from that of other cyan nephrite origins； The effect on the identification of samples are poor through cluster analysis and PCA； Artificial neural network （ANN） and Fisher discriminant analysis are used to model and identify the origin of cyan nephrite with satisfactory results； The average accurate identification ratio of training and verification set through Fisher discrimination is 100％， and so did artificial neural network （ANN）. Both methods can accurately identify the origin of cyan nephrite， and provide a technical means for qualitative analysis of cyan nephrite origin.关键词：cyan nephrite;origin identification;near-infrared spectroscopy;discriminant model;chemometrics
- 11
- |
- 2
- |
- 0
发布时间：2023-03-01 - 摘要：Continuous improvement is the basic concept of engineering education certification.Taking the course software engineering project practice for undergraduate majored in computer science as a case study， based on the collection of the final reports submitted online by the students enrolled in the class of project practice， text analysis technologies， such as context query by keywords， feature ranking， topic recognition and so on， are utilized to identify the state of achievement of students with respect to the competency goal of the precursor course object-oriented programming in Java， in order to find the continuous improvement goals. This case study provides a method， process and implementation tool based on text analysis for identifying the continuous improvement goals of programming courses.关键词：data driven teaching reform;continuous improvement;text analysis
- 22
- |
- 4
- |
- 0
发布时间：2023-03-01 - 摘要：In view of the problem that enterprises need to accurately predict the sales volume of vehicles in order to formulate a scientific and reasonable production plan in the automotive industry， taking new energy vehicles as an example， using the historical sales volume data to build a ARIMA prediction model based on traditional time series theory and an LSTM prediction model based on deep learning， and uses the Stacking method to integrate the above two models. The experimental results show that the average absolute percentage error of the combined model is 3.65%， reduced by 2.57% and 1.86% respectively，compared with ARIMA（0，1，1） model and LSTM model. This method can provide data reference for the production plan of new energy vehicles， and is applicable to other type of car.关键词：new energy vehicles;sales forecast;combined forecasting model
- 20
- |
- 3
- |
- 0
发布时间：2023-03-01 - 摘要：An improved personal credit integration classification model combining K-means clustering and rough set was proposed to solve the problem that most personal credit data have mixed data types and it is difficult to determine the initial cluster center and number of traditional K-means clustering. Firstly， the clustering degree of sample points was measured based on the density of sample space to determine the initial cluster centers， and the improved adaptive idea was introduced to dynamically adjust the number of cluster centers for K-means clustering， so as to realize the discretization of continuous data. Secondly， rough set is used for attribute reduction to get the feature subset； Finally， an integrated model based on L1-logistic regression， elastic net-logistic regression， Bayes， decision tree and neural network is constructed combining cost sensitivity to achieve effective classification of unbalanced personal credit data. Experimental results show that compared with the existing models， the proposed integrated classification model can improve G-means by 2.96% and maximum by 5.35% on average， and F-value by 3.42% and maximum by 6.83% on UCI data set.关键词：Personal credit;K-means clustering;Rough Set;Density of pattern distribution;self-adaption;Unbalanced dataset
- 32
- |
- 7
- |
- 0
发布时间：2023-03-01 - 摘要：As an important branch of computer science， intelligent science and technology is an interdisciplinary subject. It is no longer appropriate for intelligent science and technology majors to adopt the traditional training mode of computer science. The new engineering major proposes to transform and upgrade the traditional engineering major to provide opportunities for its rapid development. Taking Nanyang Normal University as the research object， this paper summarizes the incompatibility of teachers， curriculum system， personnel training objectives and personnel training modes of intelligent science and technology specialty. According to the requirements of intelligent science and technology for new engineering talents， a multi-coordinated training mode of intelligent science and technology for local governments， schools and enterprises is proposed. This model is explored from teaching resources， teaching mode， curriculum system and practice teaching. From the perspective of exploration results， the multi-coordinated innovative talent training mode of new engineering department has certain theoretical and practical significance. It can provide reference for local universities to train innovative talents in intelligent science and technology majors.关键词：new engineering;multi-coordination;intelligent science and technology;local university;talent training
- 26
- |
- 9
- |
- 0
发布时间：2023-03-01 - 摘要：Affected by the COVID-19， online teaching has become a normal teaching method， and establishing a reasonable evaluation model for teaching ability plays an important role in ensuring the quality of online teaching. In order to fit the evaluation habits of people in real life， based on formal concept analysis and hesitant fuzzy linguistic term set （HFLTS）， the hesitant fuzzy linguistic formal context is proposed， and the hesitant fuzzy linguistic concept lattice is established. On this basis， hesitant fuzzy linguistic concept lattice is used to deal with the linguistic evaluation information given by experts. The weighted similarity between HFLTSs is defined， and the application model of hesitant fuzzy linguistic rule extraction is established based on the thinner relation of hesitant fuzzy linguistic concept lattices. Finally， the method is applied to online teaching evaluation. The proposed online teaching evaluation method is more concise，has higher practical value.关键词：hesitant fuzzy linguistic;concept lattice;rule extraction;online teaching evaluation;weighted similarity
- 9
- |
- 3
- |
- 0
发布时间：2023-03-01 - 摘要：In recent years， as the focus of computational thinking education has gradually shifted to "focusing on training effects"， evaluation has become an indispensable part. However， at present， China lacks a reasonable and effective evaluation framework to support the effective promotion of computational thinking teaching practice. Therefore， draw lessons from the three-dimensional framework of computational thinking， puts forward the evaluation framework of the core elements of computational thinking according to Bloom's educational goal classification theory， analyzes and combs the evaluation strategies from the four dimensions of concept knowledge， algorithmic thinking， problem exploration and values， and puts forward specific implementation suggestions for the evaluation framework， in order to provide some reference ideas for the in-depth research and application practice of computational thinking education.关键词：computational thinking;three-dimensional framework;core elements;bloom's classification theory of educational objectives;evaluation framework
- 9
- |
- 4
- |
- 0
发布时间：2023-03-01 -
Design of A Blended Teaching Model for Python Programming Courses Based on Open Source Projects AI导读
摘要：With the continuous development of open source culture and the continuous deepening of the digitalization process， the new information-based teaching model of open source innovation combined with online teaching platforms has become a new trend for colleges and universities to carry out teaching reforms for cultivating innovative professionals.By analyzing the current situation of Python programming class， combined with the application of open source projects in actual teaching， a hybrid-teaching model of Python programming courses based on open source projects has been initially constructed.This model is committed to the cultivation of students' comprehensive abilities such as independent learning， teamwork， integration and innovation， and effectively improves the quality of curriculum teaching.关键词：mixed mode;open-source projects;Python programming- 16
- |
- 7
- |
- 0
发布时间：2023-03-01 - 摘要：With the advent of the Internet plus and 5G era， the traditional face-to-face teaching mode can no longer satisfy students' learning needs. According to the characteristics of strong abstraction in the course of data structure and algorithm， this course realizes the exploration of SPOC online and offline mixed teaching mode with students as the main body， teachers as the leading role and the combination of inside and outside the classroom through teaching design， reform in education and teaching effects. The mixed teaching mode enriches teaching means and methods， and integrates ideological and political education according to the characteristics of the course. Students' enthusiasm in classroom performance and learning effect are significantly improved.关键词：data structure and algorithm;mixed teaching;SPOC;ideological and political education
- 13
- |
- 4
- |
- 0
发布时间：2023-03-01 - 摘要：Currently， AI related courses are usually taught by teachers in accordance with the syllabus， ignoring the differences in the knowledge of students from different educational backgrounds. In view of this deficiency， this paper puts forward the AI curriculum reform plan based on the student's ability orientation， emphasizes that the students' ability to master knowledge is the orientation， around the core concept of engineering education， integrates multiple teaching methods such as group classroom discussion， experimental training links， teaching and research integration， and adopts the online and offline hybrid teaching mode to achieve the teaching goal of teaching students according to their aptitude. Students can choose appropriate learning channels according to their engineering practice ability， which will help teachers to cultivate students' innovation and practice ability. The research shows that the teaching reform of AI course based on students' ability orientation can mobilize students' enthusiasm to the greatest extent， achieve the expected teaching reform effect， and provide reference for other course reforms.关键词：artificial intelligence;students' ability orientation;course reform;practical cases;blending teaching
- 23
- |
- 11
- |
- 0
发布时间：2023-03-01 - 摘要：Hybrid teaching with deep integration of online and offline has attracted extensive attention for its flexible， timely and continuous learning form. Hybrid teaching is the development trend of college teaching and will become the new normal of education in the future. Taking the public course of C language programming as an example， sort out the advantages of Hybrid teaching and the current teaching situation of the course. In order to further improve the teaching effect， it is proposed to "systematize the content of teaching materials， personalize the network platform， three-dimensional learning resources and accurate teaching evaluation" to ensure the smooth development and implementation of hybrid teaching. After two years of teaching exploration， students' achievements have been greatly improved， talent training has achieved remarkable results， and team building has been steadily promoted. In the future， we also need to dynamically update the teaching content and realize the organic integration of knowledge transfer and educational effect under the normal condition of mixed teaching by integrating the ideological and political elements of the curriculum.关键词：C language programming;teaching reform;mixed teaching;evaluation system
- 27
- |
- 7
- |
- 0
发布时间：2023-03-01 - 摘要：Due to the fact that individual differences of students are very prominent and distinctive in C++ programming course， personalized learning is needed to be followed up. In addition， procedure evaluation data from blended learning could provide dynamic feedback to teaching implementation. In the paper， the study on stratified teaching based on blended learning is analyzed， and a stratified teaching method based on blended learning is designed and applied into practice. The results show the method proposed could effectively improve students’ learning ability and programming skills， and promote teaching satisfaction， learning interest and course average score.关键词：blended learning;stratified teaching;learning procedure evaluation;teaching mode
- 18
- |
- 5
- |
- 0
发布时间：2023-03-01 - 摘要：Analysis of student learning behavior data recorded by the intelligent teaching platform is conducive to the realization of data-driven hybrid teaching. To this end， the course access data is obtained through the Superstar platform， and the main reasons for the differences in learning effects of parallel classes are analyzed from multiple indicators in and out of class. The study found that continuous learning at ordinary times is easier to achieve good results than the pre exam assault， and the level of activity in the classroom atmosphere is not necessarily related to the learning effect. Moreover， due to the Hawthorne effect， students may deliberately beautify learning data. Therefore， teachers should comprehensively analyze the real learning status of students from the number of chapter visits， video watching rumination ratio， classroom test scores and other aspects in the teaching process. In addition， teachers should adopt targeted teaching methods according to the characteristics of the teaching class to achieve better teaching results. The research shows that it can provide reference and reference for teachers who will adopt mixed teaching methods.关键词：blended teaching;learning behavior data;data analysis;parallel classes
- 17
- |
- 5
- |
- 0
发布时间：2023-03-01 - 摘要：Aiming at the problems of abstract contents and difficult implementation of software architecture courses， propose a teaching method based on open source framework integrating cases， open source， and framework， introduces mainstream technologies from the industry， and selects contents with strong practicality and case support for teaching， considering application of software architecture to system implementation besides software architecture analysis and design. The teaching contents are organized based on the software quality attribute strategy and software architecture style， both of which have good flexibility in topic selection， resulting in a good tailorability according to different school hours. Using open source cases and teaching materials， students can understand the relevant knowledge points， have the ability to analyze software architecture based on source codes， and master the framework-based component software development method. The teaching evaluation results by students show that the proposed method can achieve good teaching effect.关键词：software architecture;open source framework;software quality attribute;software architecture style;multi-view model
- 10
- |
- 8
- |
- 0
发布时间：2023-03-01 -
Research on Construction of Ideological and Political Case Library for Software Engineering Course AI导读
摘要：In view of the problems such as inappropriate construction methods of the ideological and political case base of the software engineering course and abrupt and stiff cases of the course， propose a construction method of the ideological and political case base of the software engineering course， which is divided into three stages： preparation， construction and optimization. First of all， the four core activities of the method are discussed in detail： "improving the ideological and political quality of professional teachers"， "mining or sorting out the materials of Ideological and political cases"， "determining the integration point of Ideological and political elements and the proportion of different categories of Ideological and political elements" and "building the ideological and political case base of courses"； Then， build the case base of software engineering course. The practical results show that the ideological and political case base constructed by this method can improve the students' learning enthusiasm， and provide inspiration for similar colleges to integrate the ideological and political courses into software engineering teaching.关键词：course ideology and politics;integration points of ideological and political elements;software engineering;case library;construction research- 36
- |
- 7
- |
- 0
发布时间：2023-03-01 - 摘要：The new round of compulsory education curriculum standards proposes to vigorously develop students' core literacy. Computing thinking is not only the necessary literacy of students in the intelligent era， but also the necessary ability of students. Therefore， based on game programming to explore the appropriate teaching mode to improve the computing thinking of middle and low age students. First of all， it analyzes the function of programming teaching on improving students' computing thinking. Then， it discusses the characteristics of game programming and the feasibility of improving students' computing thinking. Finally， the teaching practice strategy of game programming is studied. The practice shows that the proposed method can improve the students' programming teaching experience and improve their computing thinking ability， which can provide reference for targeted teaching in colleges and universities.关键词：gamified programming;computational thinking;teaching practice;computer-based teaching
- 15
- |
- 3
- |
- 0
发布时间：2023-03-01 - 摘要：Acute coronary syndrome is an acute severe syndrome caused by the decrease of coronary blood flow. It is an important cause of death and long-term severe disability in patients all over the world. Predicting adverse cardiovascular events is of great significance and value for risk screening， early diagnosis and treatment of patients. Machine learning can explore new possibilities and reveal the hidden relationship in the big data of patient information statistics， which will have a positive impact on the auxiliary diagnosis and prognostic analysis of cardiovascular diseases. This paper expounds the commonly used risk scoring tools in clinic， and introduces the factors mainly dependent on patients' physiological health indicators， past medical history and other factors， as well as the ability of Cox regression model to quickly screen high-risk factors. Review the application of different machine learning models， including random forest， support vector machine and neural network， in assessing the risk of patients with acute coronary syndrome， as well as the relevant characteristics and ability in predicting long-term and short-term adverse cardiovascular events. Finally， it looks forward to the broad prospect of the application of machine learning algorithm in medical data.关键词：acute coronary syndrome;adverse cardiovascular events;machine learning;big data;auxiliary diagnosis
- 19
- |
- 6
- |
- 0
发布时间：2023-03-01 - 摘要：Event extraction is the main task of information extraction as well as one of the key technologies to construct knowledge graph. Event extraction aims to extract specific event information from unstructured natural language， which has important application value in monitoring， risk analysis， decision support and other fields. According to the different methods， the related research on event extraction is divided into three categories： data-driven， knowledge-driven and mixed extraction. Taking decision support system based on aircraft maintenance knowledge graph as an example， the application of event extraction technology in aircraft maintenance field is described in detail， and the existing problems and future development direction are summarized.关键词：event extraction;knowledge graph;natural language processing;aircraft maintenance
- 19
- |
- 7
- |
- 0
发布时间：2023-03-01 - 摘要：As a basic task of machine processing， Chinese word segmentation is one of the research hotspots in recent years. The results have a far-reaching impact on the follow-up processing tasks， and are of full research significance. Through the comprehensive analysis of the research literature on word segmentation technology in the past five years， it is clear that the follow-up research will be dominated by the fusion method based on neural network model， and further pursue more accurate and efficient word segmentation performance. In the development and application of word segmentation technology， there are also various bottlenecks restricting its performance. In addition to the traditional ambiguity and unknown words， word segmentation is now faced with new problems such as corpus scale and quality dependence and multi-domain word segmentation. The breakthrough research on these new problems will become one of the focuses of the follow-up research.关键词：Chinese word segmentation;deep learning;corpus dependence;multi-domain participle
- 20
- |
- 6
- |
- 0
发布时间：2023-03-01 - 摘要：As the word "new infrastructure construction " is put forward， education new infrastructure construction has become a hot topic in the field of education.In order to better understand the research status of the new educational infrastructure construction， summarize the related research of the new educational infrastructure construction in China， and provides reference for domestic scholars in related fields.Using content analysis and social network analysis methods， analyze the literatures related to new education infrastructure construction published from 2018 to 2021， selects 58 relevant literatures， and uses Ucinet6 and NVivo12 to conduct visual analysis.Ucinet was used to make a visual social network， and the author's cooperative research group， keyword network and literature citation were sorted out and visualized.Content analysis is carried out from four aspects： history， literature source， research content and research trend.At present， the research focus of new educational infrastructure construction lies in the application research of policy interpretation， connotation and significance， functional characteristics， promotion strategy and talent training. At present， the network connection of authors in the field is not high， and a complete research system has not yet been formed. With "Opinions" as the vane， the research focuses on six "new" directions. It is pointed out that the development of relevant standards and rules， the exploration of resource platform， the reform of education and teaching mode， and the innovation of educational technology resources will become the main research trends.关键词：new infrastructure construction for education;content analysis;social network analysis;research overview
- 8
- |
- 2
- |
- 0
发布时间：2023-03-01 - 摘要：With innovation becoming an important engine driving social and economic development， cultivating students' critical thinking has been highly valued by Education researchers at home and abroad. Using the meta-analysis method， this study analyzes the data of 31 experiments and quasi experimental reports on promoting the development of students' critical thinking in online education at home and abroad from the perspectives of online learning tools， online teaching methods， online interactive activities， experimental cycle， discipline nature and education level. The results show that online education is an effective way to promote the development of students' critical thinking， but different types of online learning tools have significant differences in promoting the development of students' critical thinking； Although there is no significant difference in the impact of different types of online interactive activities and online teaching methods on the development of students' critical thinking， there is a preference for online teaching methods， and the guiding activities do not have statistical significance in the sense of meta-analysis. Finally， the analysis results are deeply discussed， and suggestions for future research are given.关键词：critical thinking;online education;meta-analysis
- 5
- |
- 2
- |
- 0
发布时间：2023-03-01 - 摘要：As a typical application of artificial intelligence in the field of education， robot has attracted the attention of many scholars. The innovation， comprehensiveness， practicality and interdisciplinary integration of robot are similar to the concept of stem education. It is considered to be an effective way to improve students' stem discipline literacy and achieve the goal of stem Education. Foreign countries have carried out robot related research integrating stem， while domestic empirical research is still lacking. Based on this， from the perspective of stem integration， this paper combs the international robot education research from three aspects： teaching value， teaching design and teaching mode. Then， the problems existing in the practice of robot education are analyzed from teaching effect， curriculum design and teaching mode. Finally， explore the enlightenment of the research to the development of robot education in China， in order to provide reference for the practice and application of robot education in China.关键词：robotics education;STEM;empirical research;teaching practice;scientific research
- 15
- |
- 4
- |
- 0
发布时间：2023-03-01
- Technical support is provided by Beijing Founder electronics co., LTD 鄂ICP备12016542号-70
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0