Computing and Informatics (E-Journal - Institute of Informatics, SAS, Bratislava)
Not a member yet
1506 research outputs found
Sort by
Data Mining Algorithm for Web Learning Resource Information Flow Loss Based on Weighted Depth Forest
Full text linkWhen processing the lost data of web learning resource information flow, the noise in the data signal cannot be eliminated, resulting in inaccurate detection of the lost data of web learning resource information flow in the later stage. Therefore, a data mining algorithm is proposed based on weighted depth forest for web learning resource information flow loss. Based on building a brand-driven Web data acquisition model to collect data, this method uses clustering analysis technology to extract the lost data feature information of web learning resource information flow. It carries out wavelet threshold denoising on it. According to the characteristics of lost data, the lost data mining of web learning resource information flow is completed. Experimental results show that the proposed algorithm has a low error rate, high accuracy, high labour intensity, high efficiency and high performance
Solving the Two-Level Hierarchical Covering Location Problem with an Electromagnetism-Like Metaheuristic
Full text linkIn this paper, an electromagnetism-like approach (EM) for solving the two-level hierarchical covering location problem (TLHCLP) is proposed. An EM metaheuristic is a powerful algorithm for global optimization that converges rapidly to the optimum. Therefore, it has the potential to solve this type of problem since movement based on the attraction-repulsion mechanisms, combined with the proposed scaling technique, directs EM to promising search regions. The fast implementation of the objective function and local search procedure for TLHCLP additionally improves the efficiency of the overall EM system. The proposed EM approach reaches all optimal solutions in a relatively short amount of computational time. EM also obtains high-quality solutions for large-scale problem instances that are out of reach for exact methods
Quantum-Behaved Bat Algorithm Combining Convergence Factor and Self-Learning Mutation Strategies for Optimization
Full text linkQuantum-behaved Bat Algorithm (QBA) has been successfully applied as an optimal technique for dealing with a variety of optimization problems. Nevertheless, QBA suffers from similar problems as other swarm intelligent algorithms, such as poor exploration search and falling into local optima in certain conditions. Aiming at these shortcomings, an improved algorithm that combines convergence factor and gold sinusoidal self-learning mutation strategies (CGQBA) is proposed. A directional convergence factor is designed for the global position update process, it can improve the exploration search ability of the algorithm. Meanwhile, a self-learning predictive mutation mechanism is added to the algorithm. It contributes to the algorithm to jump out of the local extremum. The improved CGQBA algorithm is tested on 20 test functions with different characteristics in the numerical simulation experiments. The results and statistical tests show that CGQBA algorithm has better convergence speed, accuracy and stability. What is more, the multi-threshold image segmentation is modelled as an optimization problem, CGQBA algorithm is applied to the optimization problem to further verify the effectiveness and practicability in the real-world optimization. The results compared with three classical segmentation methods illustrate that CGQBA algorithm can effectively solve the image segmentation problem. It has a better segmentation effect and anti-noise ability
The Image Construction of News Anchors Facing Virtual Reality in the Metaverse Environment
No full textWith the continuous improvement of communication and computing technology, the technical threshold of the Metaverse (MVS) has been lowered, gradually expanding the scope of technology penetration, and attracting attention to the construction of characters in the MVS environment. In virtual reality (VR), character design and scene realizations are the foundation and guarantee of technology application. VR also provides an effective tool for the transformation of traditional media and the exploration of news communication innovation. On this basis, firstly, the introduction of VR is analyzed in the MVS environment. Secondly, based on the application status of VR, the modeling design and structure optimization of three-dimensional (3D) characters in the MVS environment are studied. Furthermore, by improving the traditional 3D modeling process, an adaptive hierarchical detail model is proposed to realize the scene modeling mechanism of the virtual environment combining image modeling and geometric modeling, so as to quickly complete the construction of the character image model of a news anchor in the scene. Finally, the model of the experimental design is simulated and constructed to test the effect of the model to construct the image of news anchors. The results reveal that the frame rate of news anchor images designed by the model can be kept at about 40 frames. The actual value of modeling details is always greater than the ideal value of modeling design. By comparing the actual modeling time of the control group and the experimental group, it can be found that the modeling time of the experimental group is shorter. Therefore, it can be concluded that the VR-based hierarchical detail model in the MVS environment designed in this research can better complete the image construction of news anchors. The research results can provide a reference and theoretical basis for the subsequent modeling work in the MVS environment and the construction of VR characters
Display Space and Distribution Visualization of Intelligent Algorithms and 3D Interactive Imaging Technology
Full text linkThe traditional two-dimensional data visualization method is difficult to accurately express and display the three-dimensional information of the data, while the 3D interactive image technology can visually display the data in the three-dimensional space. At the same time, it can also operate and edit the data through interactive methods to help people better understand and use the data. Intelligent algorithms are an important supplement to 3D interactive imaging technology, which can improve visualization and analysis accuracy through calculations, model building, and other methods, achieving more accurate data display and analysis. This article aimed to explore the application of intelligent algorithms and 3D interactive imaging technology in the field of spatial visualization. Intelligent algorithms can help analyze data features and relationships, extract data attributes, and optimize data display effects. 3D interactive imaging technology can present processed data in a virtual three-dimensional space, enhancing the interaction and visualization effect between users and data. The combination of the two can achieve better data analysis and presentation results. This article conducted simulation experiments on spatial visualization based on intelligent algorithms and 3D interactive imaging technology, and scored some experimental indicators using a scoring system. The experimental results of this article indicate that the intelligent algorithm had a time efficiency of 46.5 seconds and an accuracy of 95 % in displaying spatial and distribution visualization. The usability score was 75. The interactivity score of 3D interactive imaging technology in displaying spatial and distribution visualization was 96. The visual effect score was 96, and the usability score was 65. The combination of the two in terms of displaying spatial and distribution visualization had a time efficiency of 50.8 seconds, interactivity of 86 minutes, and accuracy of 98 %. It shows that different data visualization technologies have their own characteristics in display space and distribution visualization, and suitable technologies can be selected according to actual needs and application scenarios
SASRNet: Slimming-Assisted Deep Residual Network for Image Steganalysis
Full text linkExisting deep-learning-based image steganalysis networks have problems such as large model sizes, significant runtime memory usage, and extensive computational operations, which hinder their deployment in many practical applications. To address these challenges, we applied model compression techniques to image steganalysis and designed a model called SASRNet, a slimming-assisted steganalysis residual network. We observed that the trainable scale factor of BN (batch normalization) layer in steganalysis network can be used as channel scaling factor for pruning. The channel-level sparsity of convolutional layers is enhanced by imposing L1 regularization on channel scaling factors and pruning less informative feature channels. With the goal of balancing performance and efficiency, the iterative algorithm is used to further compress the network to obtain a slimming- steganalysis detector. In contrast to many existing methods, our proposed method can be directly applied to steganalysis network architectures by introducing a minimal overhead to the training process. We have conducted extensive experiments on BOSSBase+BOWS2 dataset. Experiments show that, compared to the original steganalysis model, this method can achieve comparable performance with less than 5 % of the parameters, validating the feasibility and practicality of the new model
Prediction of Stress Level from Speech – from Database to Regressor
Full text linkThe term stress can designate a number of situations and affective reactions. This work focuses on the immediate stress reaction caused by, for example, threat, danger, fear, or great concern. Could measuring stress from speech be a viable fast and non-invasive method? The article describes the development of a system predicting stress from voice – from the creation of the database, and preparation of the training data to the design and tests of the regressor. StressDat, an acted database of speech under stress in Slovak, was designed. After publishing the methodology during its development in [1], this work describes the final form, annotation, and basic acoustic analyses of the data. The utterances presenting various stress-inducing scenarios were acted at three intended stress levels. The annotators used a "stress thermometer" to rate the perceived stress in the utterance on a scale from 0 to 100. Thus, data with a resolution suitable for training the regressor was obtained. Several regressors were trained, tested and compared. On the test-set, the stress estimation works well (R square = 0.72, Concordance Correlation Coefficient = 0.83) but practical application will require much larger volumes of specific training data. StressDat was made publicly available
Does a Robot's Gaze Behavior Affect Entrainment in HRI?
Full text linkSpeakers tend to engage in adaptive behavior, known as entrainment, when they reuse their partner's linguistic representations, including lexical, acoustic prosodic, semantic, or syntactic structures during a conversation. Studies have explored the relationship between entrainment and social factors such as likeability, task success, and rapport. Still, limited research has investigated the relationship between entrainment and gaze. To address this gap, we conducted a within-subjects user study (N = 33) to test if gaze behavior of a robotic head affects entrainment of subjects toward the robot on four linguistic dimensions: lexical, syntactic, semantic, and acoustic-prosodic. Our results show that participants entrain more on lexical and acoustic-prosodic features when the robot exhibits well-timed gaze aversions similar to the ones observed in human gaze behavior, as compared to when the robot keeps staring at participants constantly. Our results support the predictions of the computers as social actors (CASA) model and suggest that implementing well-timed gaze aversion behavior in a robot can lead to speech entrainment in human-robot interactions
Metaheuristic for Solving the Delivery Man Problem with Drone
Full text linkDelivery Man Problem with Drone (DMPD) is a variant of Delivery Man Problem (DMP). The objective of DMP is to minimize the sum of customers' waiting times. In DMP, there is only a truck to deliver materials to customers while the delivery is completed by collaboration between truck and drone in DMPD. Using a drone is useful when a truck cannot reach some customers in particular circumstances such as narrow roads or natural disasters. For NP-hard problems, metaheuristic is a natural approach to solve medium to large-sized instances. In this paper, a metaheuristic algorithm is proposed. Initially, a solution without drone is created. Then, it is an input of split procedure to convert DMP-solution into DMPD-solution. After that, it is improved by the combination of Variable Neighborhood Search (VNS) and Tabu Search (TS). To explore a new solution space, diversification is applied. The proposed algorithm balances diversification and intensification to prevent the search from local optima. The experimental simulations show that the proposed algorithm reaches good solutions fast, even for large instances
Novel Technique of Healthcare Record Indexing and Recommendation Based on Trending Queries in Social Media
Full text linkRecommendation of services and applications based on user-data analytics is the most common approach to understanding user requirements. In this article, a novel technique for user recommendation is proposed and validated. The technique uses a Twitter Application Programming Interface (API) handle-based dataset for evaluating and computing the recommendations. The technique uses an open platform Graphical User Interface (GUI) for keyword categorization and building a reliable support system for query analysis. API driven queries from Twitter are cross-validated with labeling techniques and trending hashtags. Typically, the defined tweets are validated to build a Healthcare Record Indexing (HRI) data structure. The HRI is used to support the decision-making and recommendation of services of various healthcare applications and tweets. The technique has trained 750 datasets of categorized clusters with 150 000 tweets (dynamic) datasets from Twitter API. The technique has performed 92.68 % in accuracy and 91.72 % in sensitivity of given datasets