Digital Library[ Search Result ]
Optimizing Throughput Prediction Models Based on Feature Category Contribution in 4G/5G Network Environments
http://doi.org/10.5626/JOK.2024.51.11.961
The acceleration in 5G technology adoption due to increased network data consumption and limitations of 4G has led to the establishment of a heterogeneous network environment comprising both 4G and limited 5G. Consequently, this highlights the importance of throughput prediction for network service quality (QoS) and resource optimization. Traditional throughput prediction research mainly relies on the use of single attributes or extraction of attributes through correlation analysis. However, these approaches have limitations, including potential exclusion of variables with nonlinear relationships with arbitrariness and inconsistency of correlation coefficient thresholds. To overcome these limitations, this paper proposed a new approach based on Feature Importance. This method could calculate the relative importance of features used in the network and assign contribution scores to attribute categories. By utilizing these scores, throughput prediction was enhanced. This approach was applied and tested on four open network datasets. Experiments demonstrated that the proposed method successfully derived an optimal category combination for throughput prediction, reduced model complexity, and improved prediction accuracy compared to using all categories.
Lightweight Temporal Segment Network for Video Scene Understanding: Validation in Driver Assault Detection
Juneyong Lee, Joon Kim, Junhui Park, Jongho Jo, Ikbeom Jang
http://doi.org/10.5626/JOK.2024.51.11.987
"The number of driver assaults in transportation such as taxis and buses has been increasing over the past few years. It can be especially difficult to respond quickly to assaults on drivers by drunks late at night. To address this issue, our research team proposed a lightweight CNN-based Temporal Segment Network (TSN) model that could detect driver assaults by passengers in real time. The TSN model efficiently processes videos by sampling a small number of image frames and divides videos into two streams for learning: one for spatial information processing and the other for temporal information processing. Convolutional neural networks are employed in each stream. In this research, we applied a lightweight CNN architecture, MobileOne, significantly reducing the model size while demonstrating improved accuracy even with limited computing resources. The model is expected to contribute to rapid response and prevention of hazardous situations for drivers when it is integrated into vehicular driver monitoring systems."
Cardiovascular Disease Prediction using Single-Lead ECG Data
Chaeyoon Park, Gihun Joo, Suhwan Ji, Junbeom Park, Junho Baek, Hyeonseung Im
http://doi.org/10.5626/JOK.2024.51.10.928
The most representative approach to diagnosing cardiovascular disease is to analyze electrocardiogram (ECG), and most ECG data measured in hospitals consist of 12 leads. However, wearable healthcare devices usually measure only single-lead ECG, which has limitations in diagnosing cardiovascular disease. Therefore, in this paper, we conducted a study to predict common cardiovascular diseases such as atrial fibrillation (AF), left bundle branch block (LBBB), and right bundle branch block (RBBB) using a single lead that could be measured with a wearable healthcare device. For experiments, we used a convolutional neural network model and measured its performance using various leads in terms of AUC and F1-score. For AF, LBBB, and RBBB, average AUC values were 0.966, 0.971, and 0.965, respectively, and average F1-scores were 0.867, 0.816, and 0.848, respectively. These experimental results confirm the possibility of diagnosing cardiovascular disease using only a single lead ECG that can be obtained with wearable healthcare devices.
A Comparative Study on Server Allocation Optimization Algorithms for Accelerating Parallel Training of Large Language Models
Jinkyu Yim, Yerim Choi, Jinho Lee
http://doi.org/10.5626/JOK.2024.51.9.783
As large-scale language models (LLMs) come to be increasingly utilized in various fields, there is an increasing demand to develop models with higher performance. Significant computational power and memory capacity will be needed to train such models. Therefore, researchers have used 3D parallelization methodology for large-scale language model learning on numerous servers equipped with GPUs. However, 3D parallelization requires frequent large-scale data transfers between servers, which bottlenecks the overall training time. To address this, prior studies have proposed a methodology that identifies non-uniform cluster network conditions in advance and arranges servers and GPUs in an optimized parallel configuration. The existing methods of this type use the classical optimization algorithm SA (Simulated Annealing) for mapping. In this paper, we apply genetic algorithms as well as SAT(satisfiability) algorithms to the problem, and compare and analyze the performance of each algorithm under various experimental environments.
Photovoltaic Power Forecasting Scheme Based on Graph Neural Networks through Long- and Short-Term Time Pattern Learning
Jaeseung Lee, Sungwoo Park, Jaeuk Moon, Eenjun Hwang
http://doi.org/10.5626/JOK.2024.51.8.690
As the use of solar energy has become increasingly common in recent years, there has been active research in predicting the amount of photovoltaic power generation to improve the efficiency of solar energy. In this context, photovoltaic power forecasting models based on graph neural networks have been presented, going beyond existing deep learning models. These models enhance prediction accuracy by learning the interactions between regions. Specifically, they consider how the amount of photovoltaic power in a specific region is affected by the climate conditions of adjacent regions and the time pattern of photovoltaic power generation. However, existing models mainly rely on a fixed graph structure, making it difficult to capture temporal and spatial interactions. In this paper, we propose a graph neural networks-based photovoltaic power forecasting scheme that takes into account both long-term and short-term time patterns of regional photovoltaic power generation data. We then incorporate these patterns into the learning process to establish correlations between regions. Compared to other graph neural networks-based prediction models, our proposed scheme achieved a performance improvement of up to 7.49% based on the RRSE, demonstrating its superiority.
Continual Learning using Memory-Efficient Parameter Generation
Hyung-Wook Lim, Han-Eol Kang, Dong-Wan Choi
http://doi.org/10.5626/JOK.2024.51.8.747
Continual Learning with Parameter Generation shows remarkable stability in retaining knowledge from previous tasks. However, it suffers from a gradual decline in parameter generation performance due to its lack of adaptability to new tasks. Furthermore, the difficulty in predetermining the optimal size of the parameter generation model (meta-model) can lead to memory efficiency issues. To address these limitations, this paper proposed two novel techniques. Firstly, the Chunk Save & Replay (CSR) technique selectively stored and replayed vulnerable parts of the generative neural network, maintaining diversity in the parameter generation model while efficiently utilizing memory. Secondly, the Automatically Growing GAN (AG-GAN) technique automatically expanded the memory of the parameter generation model based on learning tasks, enabling effective memory utilization in resource-constrained environments. Experimental results demonstrated that these proposed techniques significantly reduced memory usage while minimizing performance degradation. Moreover, their ability to recover from deteriorated network performance was observed. This research presents new approaches to overcoming limitations of parameter generation-based continual learning, facilitating the implementation of more effective and efficient continual learning systems.
A Survey of Advantages of Self-Supervised Learning Models in Visual Recognition Tasks
Euihyun Yoon, Hyunjong Lee, Donggeon Kim, Joochan Park, Jinkyu Kim, Jaekoo Lee
http://doi.org/10.5626/JOK.2024.51.7.609
Recently, the field of teacher-based artificial intelligence (AI) has been rapidly advancing. However, teacher-based learning relies on datasets with specified correct answers, which can increase the cost of obtaining these correct answers. To address this issue, self-supervised learning, which can learn general features of photos without needing correct answers, is being researched. In this paper, various self-supervised learning models were classified based on their learning methods and backbone networks. Their strengths, weaknesses, and performances were then compared and analyzed. Photo classification tasks were used for performance comparison. For comparing the performance of transfer learning, detailed prediction tasks were also compared and analyzed. As a result, models that only used positive pairs achieved higher performance by minimizing noise than models that used both positive and negative pairs. Furthermore, for fine-grained predictions, methods such as masking images for learning or utilizing multi-stage models achieved higher performance by additionally learning regional information.
Improving Prediction of Chronic Hepatitis B Treatment Response Using Molecular Embedding
Jihyeon Song, Soon Sun Kim, Ji Eun Han, Hyo Jung Cho, Jae Youn Cheong, Charmgil Hong
http://doi.org/10.5626/JOK.2024.51.7.627
Chronic hepatitis B patients with no timely treatment are at a high risk of developing complications such as liver cirrhosis and hepatocellular carcinoma (liver cancer). As a result, various antiviral agents for hepatitis B have been developed, and due to the different components of these antiviral agents, there can be variations in treatment responses among patients. Therefore, selecting the appropriate medication that leads to a favorable treatment response is considered crucial. In this study, in addition to the patient's blood test results and electronic medical records indicating drug prescriptions, information about components of the hepatitis B antiviral agents was incorporated for learning. The aim was to enhance the prediction performance of treatment responses one year after chronic hepatitis B patients' treatment. Molecular embedding of the antiviral agents included both fixed molecular embedding and those generated through an end-to-end structure utilizing a graph neural network model. By comparing with the baseline model, drug molecule embedding was confirmed to contribute to improving performance.
Prediction of Cancer Prognosis Using Patient-Specific Cancer Driver Gene Information
http://doi.org/10.5626/JOK.2024.51.6.574
Accurate prediction of cancer prognosis is crucial for effective treatment. Consequently, numerous studies on cancer prognosis have been conducted, with recent research leveraging various machine learning techniques such as deep learning. In this paper, we first constructed patient-specific gene networks for each patient, then selected patient-specific cancer driver genes, considering the heterogeneity of cancer. We propose a deep neural architecture that can predict the prognosis more accurately using patient-specific cancer driver gene information. When our method was applied to gene expression data for 11 types of cancer, it demonstrated a significantly higher prediction accuracy compared to the existing methods.
aRFS+: A New Flow Steering Scheme for High Network Performance
http://doi.org/10.5626/JOK.2024.51.5.397
Recent studies indicate that a significant portion of central processing unit (CPU) usage in network stack processing is attributed to the transfer of data between kernel and user spaces. Direct Cache Access (DCA) has been recognized to enhance data copy efficiency by allowing applications to perform data copy operations utilizing L3 caches. However, current flow steering mechanisms lack awareness of caches; they often employ random selection methods or allocate processing tasks to cores based on the location of corresponding applications subsequently resulting in suboptimal throughput. To address this issue, in this paper, we propose a novel flow steering scheme named aRFS+. The three key ideas of aRFS+ are as follows. First, we dynamically allocated network applications to the DCA-capable NUMA node, enabling them to exploit DCA advantages during data copy operations. Second, we decouple application cores from network processing cores to maximize the benefits of multicore environments. Incoming packets are steered to a CPU distinct from the application core but situated within the same NUMA node. Third, we introduce an optimization technique that significantly mitigates the overhead associated with memory management. Our experimental evaluations demonstrated that aRFS+ substantially improved the overall throughput, with an enhancement of up to 60% compared to existing schemes.
Search
Journal of KIISE
- ISSN : 2383-630X(Print)
- ISSN : 2383-6296(Electronic)
- KCI Accredited Journal
Editorial Office
- Tel. +82-2-588-9240
- Fax. +82-2-521-1352
- E-mail. chwoo@kiise.or.kr