As a result of ongoing advances in artificial intelligence technology, the potential for learning analysis in teaching evaluation and educational data mining is gradually being recognized. In classrooms, artificial intelligence technology can help to enable automated student behavior analysis, so that teachers can effectively and intuitively grasp students’ learning behavior engagement; the technology, moreover, can provide data to support subsequent improvements in learning design and implementation of teaching interventions. The main scope of the research is as follows: Construct a classroom student behavior dataset that provides a basis for subsequent research; Propose a behavior detection method and a set of feasible, high-precision behavior recognition models. Based on the global features of the human posture extracted from the Openpose algorithm and the local features of the interactive objects extracted by the YOLO v3 algorithm, student behavior can be identified and analyzed to help improve recognition accuracy; Improve the model structure, compress and optimize the model, and reduce the consumption of computing power and time. Four behaviors closely related to the state of learning engagement: listening, turning sideways, bowing, and raising hands are recognized. The accuracy of the detection and recognition method on the verification set achieves 95.45%. The recognition speed and accuracy of common behaviors, such as playing with mobile phones and writing, are greatly improved compared to the original model.
Network traffic anomaly detection based on feature selection has attracted great research interest. Most existing schemes detect anomalies by reducing the dimensionality of traffic data, but ignore the correlation between data features; this results in inefficient detection of anomaly traffic. In order to effectively identify various types of attacks, a model based on a self-attentive mechanism is proposed to learn the correlation between multiple features of network traffic data. Then, a novel multi-feature anomalous traffic detection and classification model is designed, which analyzes the correlation between multiple features of the anomalous traffic data and subsequently identifies anomalous network traffic. Experimental results show that, compared to two benchmark methods, the proposed technique increased the accuracy of anomaly detection and classification by a maximum of 1.65% and reduced the false alarm rate by 1.1%.
This paper aims to address the challenge of seeking an optimal safe path for a UAV (unmanned aerial vehicle) from an initial position to a target position, while avoiding all obstacles in a three-dimensional environment. An improved APF (artificial potential field) method combined with the regular hexagon guidance method is proposed to solve unreachable and local minimum problems near obstacles as observed with traditional artificial potential field methods. First, we add a distance correction factor to the repulsive potential field function to solve problems associated with unreachable targets. Then, a regular hexagon-guided method is proposed to improve the local minimum problem. This method can judge the environment when the UAV is trapped in a local minimum point or trap area and select the appropriate planning method to guide the UAV to escape from the local minimum area. Then, 3D modeling and simulation were carried out via Matlab, taking into account a variety of scenes involving complex obstacles. The results show that this method has good feasibility and effectiveness in real-time path planning of UAVs. Lastly, we demonstrate the performance of the proposed method in a real environment, and the experimental results show that the proposed method can effectively avoid obstacles and find the optimal path.
With the emergence of low-latency applications such as driverless cars, online gaming, and virtual reality, it is becoming increasingly difficult to meet users’ demands for service quality using the traditional centralized mobile cloud computing model. In order to make up for the shortages of cloud computing, mobile edge computing came into being, which provides users with computing and storage resources by migrating computing tasks to network edge servers through computation offloading. However, most of the existing work processes only consider single-objective performance optimization of delay or energy consumption, and do not consider the balanced optimization of delay and energy consumption. Therefore, in order to reduce task delay and equipment energy consumption, a multi-user joint computation offloading and resource allocation strategy is proposed. In this strategy, the Lagrange multiplier method is used to obtain the optimal allocation of computing resources for a given offloading decision. Then, a computation offloading algorithm based on a greedy algorithm is proposed to obtain the optimal offloading decision; the final solution is obtained through continuous iteration. Experimental results show that, compared with the benchmark algorithm, the proposed algorithm can reduce system costs by up to 40%.
Adaptive learning is an educational method that uses computer algorithms to coordinate interaction with learners, and provides customized learning resources and learning activities to address the unique needs of each learner. With the impact of COVID-19, adaptive learning has become increasingly important. One of the challenges with adaptive learning is how to provide personalized learning resources for learners—i.e., how to generate personalized recommendation for learners from a large set of learning resources. Existing methodologies mainly generate recommendations based on a learner’s knowledge level; however, this approach has some limitations. Firstly, when assessing a learner’s knowledge level, learners’ forgetting phenomenon has to date not been well modeled. Secondly, recommendations are generated separately from knowledge tracing tasks, ignoring the interconnectedness between these aspects. In addition, learners’ preferences for the type of learning resources and learning strategies is normally ignored if the knowledge level alone is used. To solve the aforementioned problems, this paper proposes a knowledge and personality incorporated multi-task learning framework (KPM) to boost course recommendations (i.e., the above-mentioned learning resources); the proposed method regards an enhanced knowledge tracing task (EKTT) as an auxiliary task to assist the primary course recommendation task (CRT). Specifically, using EKTT, we design a personalized forgetting controller to enhance the deep knowledge tracing model for accurately assessing a learner’s knowledge level. With CRT, we combine the learner’s knowledge level and sequential behavior with their personality adapted to the specific context to obtain learner’s profile; this data is subsequently used to generate a course recommendation list. Experimental results on real-world educational datasets demonstrate the superiority of our proposed method in terms of hit ratio (HR), normalized discounted cumulative gain (NDCG), and precision, indicating that our method can generate more personalized recommendations.
Image stitching technology is one of the key technologies in the application of large-field microscopic digital images. The existing traditional image stitching method is to stitch in a fixed order after image registration, and once there is an error, it will be accumulated along a fixed path, thereby causing problems such as misalignment of subsequent images. In this study, through experimental analysis, a method for optimizing the stitching path of the large-field image was proposed, which greatly optimized the problems caused by error accumulation and registration failure, and effectively improved the stitching quality of the large-field microscopic digital image. This method can be used not only for the stitching of large-field microscopic images, but also for other types of stitching.
Battlefield environments are combat spaces that contain geographic elements such as terrain and roads. Road modeling and simulation is an important part of battlefield simulation and plays a key role in complex combat decision-making. Traditional road modeling is unable to handle the complex terrain conditions present in the field; hence, this paper proposes road modeling and simulation method for field environments. In particular, in order to support road modeling and simulation of complex terrain environments, road construction designs oriented to typical battlefield environments are proposed. This method divides the road network into different sub-models according to their characteristics and models them separately, improving the demand for realism in battlefield simulation. Then, the proposed method uses OpenStreetMap geographic information data to drive road network construction. The model offers real-time, high accuracy road information content and complete classification that can meet the needs of military operations and modeling simulations for typical battlefield environments. Secondly, using terrain elevation data, road construction rules, and other auxiliary information, the road height is adjusted to adapt to the complex terrain conditions of the battlefield and possible multi-level road network structures. Lastly, the introduction of a $ {G}^{2} $ continuous Hermite interpolation spline can flexibly represent the center line of the road and improves the reusability of the road model through grid deformation. Experiments show that the proposed simulation method can more reliably restore the real details of a road network to effectively fit complex terrain and improve the reusability of road models. Finally, it provides a feasible analysis angle and modeling method for researching geographic elements in battlefield environments.
The advanced capabilities of artificial intelligence (AI) have been widely used to process large volumes of data in real-time for achieving rapid response. In contrast, conventional methods for deploying various AI-based applications can result in substantial computational and communication overhead. To solve this problem, a deep model Edge-Cloud collaborative acceleration mechanism based on network compression and partitioning technology is proposed. This technology can compress and partition deep neural networks (DNN), and deploy artificial intelligence models in practical applications in the form of an Edge-Cloud collaboration for rapid response. As a first step, the proposed method compresses the neural network to reduce the execution latency required and generates a new layer that can be used as a candidate partition point. It then trains a series of prediction models to find the best partitioning point and partitions the compressed neural network model into two parts. The two parts obtained are deployed in the edge device and the cloud server, respectively, and these two parts can collaborate to minimize the overall latency. Experimental results show that, compared with four benchmarking methods, the proposed scheme can reduce the total delay of the depth model by more than 70%.
By combining information extraction technology, data matching technology, and one-time manual processing, mathematical subjective questions can be transformed into tree-shaped multiple-choice questions. In this study, an automatic correction system for college mathematics assignments was developed by combining modern information and network technology; the system was subsequently trialed in the teaching of entry-level college mathematics courses. The proposed system solves bottlenecks related to automatic grading of subjective mathematics questions, including multiple-choice questions, fill-in-the-blank questions, judgment questions, quiz questions, calculation questions, and proof questions. The system can correct routine exercises for mathematics courses of primary schools, middle schools, and universities so as to achieve more efficient completion. The idea, furthermore, can be applied to various aspects of mathematics teaching, such as previews before class, classroom exercises, reviews after class, preparations for examinations, online examinations, etc. The electronic data collected in the process of automatic correction can subsequently be used for data analysis, teaching guidance, teaching research, and the construction of educational informatization.
With the rapid development of the Internet of Things, embedded hardware products face great challenges in data security. The AES (Advanced Encryption Standard) algorithm has the advantages of strong attack resistance, fast operation speed and flexible block length in the field of data encryption and decryption. The speed of this algorithm on microcontroller platforms is far inferior to general-purpose CPUs (Central Processing Units) which have an extended instruction set for AES encryption. To solve this problem, a speed optimized AES algorithm in CTR (Counter) mode based on the Cortex-M4 core instruction set is implemented using assembly language. The kernel’s unique barrel shifter and three-stage pipeline are used to optimize the round transformation of the algorithm, and the number of instruction cycles is reduced. Testing on an FRDM-K82F development board shows that the assembly optimization of the algorithm is substantially more efficient than the code implemented using the C language, and it offers more advantages in both cost and power consumption compared to hardware encryption based on the coprocessor.
With the continuous development of Internet technology, network security is garnering increasing attention. Network anomalous traffic detection can provide an effective guarantee for blocking network attacks. However, to accurately detect anomalous traffic in a network, analyzing large volumes of data is usually required. Analyzing this data not only consumes substantial computational resources and reduces real-time detection capability, but it may also reduce the overall accuracy of detection. To solve these problems, we propose a network anomaly traffic detection method based on ensemble feature selection. Specifically, we use five different feature selection algorithms to design a voting mechanism for selecting feature subsets. Three different machine learning algorithms (Naive Bayesian, Decision Tree, XGBoost) are used to evaluate the feature selection algorithm, and the best algorithm is selected to detect abnormal network traffic. The experimental results show that the runtime of the proposed method is 84.38% less than the original data set on the optimal feature subset selected by the proposed approach, and the average accuracy is 16.93% higher than that of the single feature selection algorithm.
The concept of knowledge tracking involves tracking changes in a student’s knowledge level based on historical question records and other auxiliary information, and predicting the result of a student’s subsequent answer to a question. Since the performance of existing neural network knowledge tracking models needs to be improved, this paper proposes a deep residual network based on a stacked gated recurrent unit (GRU) network named the stacked-gated recurrent unit-residual (S-GRU-R) network. The proposed solution aims to address over-fitting caused by too many parameters in a long short-term memory (LSTM) network; hence, the solution uses a GRU instead of LSTM to learn information on the sequence of questions. The use of stacked GRU can expand sequence learning capacity, and the use of residual connections can reduce the difficulty of model training. Experiments on the Statics2011 data set were completed using S-GRU-R, and AUC (area under the curve) and F1-score were used as evaluation functions. The results showed that S-GRU-R surpassed other similar recurrent neural network models in these two indicators.
Neural architecture search algorithms aim to find more efficient neural network structures in a huge neural network structure space using computer heuristic search instead of manual search. Previous studies have addressed the problem of inefficient and time-consuming search for early neural network structures by introducing various constraints on the search space. While constraints on the search space can improve and stabilize the performance of the model, they ignore potentially efficient model structures. Hence, in this study, we constructed a recursive model search space that focuses more on the macroscopic structure of neural networks. We proposed a neural architecture search algorithm that explores this search space through a step-by-step incremental search approach. Experiments showed that the algorithm can efficiently perform neural architecture search tasks in complex search spaces, but still fell slightly short of the latest constrained search space-based neural architecture search algorithms.
The identification and classification of garment patterns are important technologies for intelligent clothing production and management. This paper proposes a method to convert garment patterns into graphic data and subsequently proposes a lightweight graph neural network GPC-GCN (Garment Pattern Classification Graph Convolutional Network) that can process this graphic data. The proposed graph data modeling method can not only maintain information on the shape of each component in the garment pattern but also deal with the arbitrariness of the position of components in garment patterns. Experiments show that the proposed graph neural network GPC-GCN achieves a better result for the classification of garment patterns compared to convolutional neural networks and graph convolutional networks.
Traditional virtual terrain modeling commonly uses a procedural generation method based on manual design, which cannot be used for competent simulation modeling tasks that need to restore real environments, such as in military applications. In this paper, we proposed a landscape simulation modeling method based on remote sensing images. The core of our proposed method is a landscape blended texture generation network (LBTG-Net); this method uses a blended texture generator (BTG) to generate landscape blended textures with the supervision of a style discriminator (SD) and multi-stage classification loss. Then, we procedurally build the complete virtual environment based on the blended texture generated by LBTG-Net. Our method has two main features: (1) accurate land-cover classification ability of remote sensing image inputs; and (2) high quality landscape blended texture outputs to guarantee virtual landscape modeling quality. We used multispectral image data from the Sentinel-2 satellite as the experimental dataset. The experimental results showed that our method offered high performance under mainstream land-cover classification evaluating indicators and can accurately reproduce the environmental distribution of input remote sensing images while completing high-quality virtual terrain simulation modeling.
With the rapid development of smart grids, the construction of new digital infrastructure has become one of the core businesses of power companies. Power companies’ governance and intelligent analytical capabilities enable opportunities for business model innovation, such as platform operation and value-added data realization. In the context of power digitization and intelligent governance, we use the robust random cut forest in this paper for transformer loss data anomaly intelligence detection. The algorithm divides sample points by random cutting to construct a random cut forest structure model by inserting and removing sample points in the structure; the anomaly score of a sample point is then given by the influence of complexity. This method is suitable for anomaly detection on real-time loss data and offers a high degree of credibility, effectiveness, and efficiency. An experiment of anomaly detection on real transformer loss data shows that the method is efficient and flexible. The accuracy, recall, and efficiency of the proposed method, moreover, is substantially better than alternatives.
Existing methods of key points matching were invented for grayscale images and are not suitable for high resolution images. Mural images typically have very high resolution, and there may be areas with the same gray textures and different colors. For this special kind of image, this paper proposes a high-speed algorithm of key points matching for high-resolution mural images (NeoKPM for short). NeoKPM has two main innovations: (1) first, the homography matrix of rough registration for the original image is obtained by downsampling the image, which substantially reduces the time required for key points matching; (2) second, a feature descriptor based on gray and color invariants is proposed, which can distinguish different colors of texture with the same gray level, thereby improving the correctness of key points matching. In this paper, the performance of the NeoKPM algorithm is tested on a real mural image library. The experimental results show that on mural images with a resolution of 80 million pixels, the number of correct matching points per pair of images is nearly 100 000 points higher than that of the SIFT (Scale Invariant Feature Transform) algorithm, the processing speed of key points matching is more than 20 times faster than that of the SIFT algorithm, and the average error of dual images based on a single pixel of the image is less than 0.04 pixels.
With the development and iteration of financial technology(FinTech) software programs, the size of test suites will gradually increase, which may introduce inherent redundancy. In order to effectively quantify test redundancy, a test redundancy evaluation metric called MVI (Most Valuable Item) is proposed in this study. To verify the validity of the MVI metric, the MVIR (Most Valuable Item Reduction) test case reduction algorithm is proposed. Experimental results show that the MVIR can achieve a test case reduction ratio of more than 89.88% assuming the test performance loss is less than 9.20%, this demonstrates that the MVI metric is valid.
In recent years, non-volatile memory (NVM) has developed rapidly. Its advantages, among others, include: persistence, large capacity, low latency, byte addressing, high density, and low energy consumption — all of which have impacted current database system architecture. SQLite is a lightweight relational database widely used in embedded fields such as mobile platforms. It operates as a serverless, zero-configuration, transactional SQL database engine. It maintains a cache for each connection, which results in problems with large space overhead and data consistency detection. At the same time, it adopts a relatively simple serialized single-write transaction execution method and page-based logging, which offers low performance and write amplification in the journal mode and a storage space requirement in the WAL mode during execution. In order to address the above challenges, a new scheme of SQLite Cache based on non-volatile memory, SQLite-CC (Copy Cache), is constructed, which fully considers the hardware characteristics of non-volatile memory and ensures the atomicity of transactions using a CC-manager and by adding an updated page index to ensure the consistency of database files and cache. Benchmarking tests show that it can achieve the same concurrency performance as SQLite-WAL mode. Compared with the rollback mode, it improves the execution performance of transactions by 3 times, reduces latency by 40%, and effectively solves the issue of write amplification on disks.
Using double-layer networks, we constructed a coupled propagation model (Noisy Voter - Susceptible-Infected-Recovery) with different time evolution scales. This coupled spreading process can be characterized by numerical analysis method of microscopic Markov chain theory. We verified the accuracy of the proposed numerical analysis method using a large number of Monte Carlo simulation experiments. We found a crossover phenomenon of the phase transition type in the coupled model. Specifically, when the noise in the opinion formation process is relatively small, the information propagation scale and the proportion of positive opinions change discontinuously with the information transmission rate. At the same time, the hysteresis loop and bistability phenomenon appear, in which the phenomenon of global consensus can be observed. When the noise is large, the order parameters of these two dynamic processes vary continuously with the transmission rate.
中国综合性科技类核心期刊(北大核心)
