• 전자통신동향분석
• /
• 제38권6호
• /
• pp.1-11
• /
• 2023

Large language models seem promising for handling reasoning problems, but their underlying solving mechanisms remain unclear. Large language models will establish a new paradigm in artificial intelligence and the society as a whole. However, a major challenge of large language models is the massive resources required for training and operation. To address this issue, researchers are actively exploring compact large language models that retain the capabilities of large language models while notably reducing the model size. These research efforts are mainly focused on improving pretraining, instruction tuning, and alignment. On the other hand, chain-of-thought prompting is a technique aimed at enhancing the reasoning ability of large language models. It provides an answer through a series of intermediate reasoning steps when given a problem. By guiding the model through a multistep problem-solving process, chain-of-thought prompting may improve the model reasoning skills. Mathematical reasoning, which is a fundamental aspect of human intelligence, has played a crucial role in advancing large language models toward human-level performance. As a result, mathematical reasoning is being widely explored in the context of large language models. This type of research extends to various domains such as geometry problem solving, tabular mathematical reasoning, visual question answering, and other areas.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 1998년도 봄 학술발표회 논문집
• /
• pp.291-300
• /
• 1998

This paper demonstrates how ambient vibration measurements at a limited number of locations can be effectively utilized to estimate parameters of a finite element model of a large-scale structural system involving a large number of elements. System identification using ambient vibration measurements presents a challenge requiring the use of special identification techniques, which ran deal with very small magnitudes of ambient vibration contaminated by noise without the knowledge of input farces. In the present study, the modal parameters such as natural frequencies, damping ratios, and mode shapes of the structural system were estimated by means of appropriate system identification techniques including the random decrement method. Moreover, estimation of parameters such as the stiffness matrix of the finite element model from the system response measured by a limited number of sensors is another challenge. In this study, the system stiffness matrix was estimated by using the quadratic optimization involving the computed and measured modal strain energy of the system, with the aid of a sensitivity relationship between each element stiffness and the modal parameters established by the second order inverse modal perturbation theory. The finite element models thus identified represent the actual structural system very well, as their calculated dynamic characteristics satisfactorily matched the observed ones from the ambient vibration test performed on a large-scale structural system subjected primarily to ambient wind excitations. The dynamic models identified by this study will be used for design of an active mass damper system to be installed on this structure fer suppressing its wind vibration.

• Nuclear Engineering and Technology
• /
• 제46권1호
• /
• pp.11-26
• /
• 2014

The analyses carried out within the Seismic Probabilistic Risk Assessments (SPRAs) of Nuclear Power Plants (NPPs) are affected by significant aleatory and epistemic uncertainties. These uncertainties have to be represented and quantified coherently with the data, information and knowledge available, to provide reasonable assurance that related decisions can be taken robustly and with confidence. The amount of data, information and knowledge available for seismic risk assessment is typically limited, so that the analysis must strongly rely on expert judgments. In this paper, a Dempster-Shafer Theory (DST) framework for handling uncertainties in NPP SPRAs is proposed and applied to an example case study. The main contributions of this paper are two: (i) applying the complete DST framework to SPRA models, showing how to build the Dempster-Shafer structures of the uncertainty parameters based on industry generic data, and (ii) embedding Bayesian updating based on plant specific data into the framework. The results of the application to a case study show that the approach is feasible and effective in (i) describing and jointly propagating aleatory and epistemic uncertainties in SPRA models and (ii) providing 'conservative' bounds on the safety quantities of interest (i.e. Core Damage Frequency, CDF) that reflect the (limited) state of knowledge of the experts about the system of interest.

• 로봇학회논문지
• /
• 제19권1호
• /
• pp.106-116
• /
• 2024

Ensuring robust 3D object detection is a core challenge for autonomous driving systems operating in urban environments. To tackle this issue, various 3D representation, including point cloud, voxels, and pillars, have been widely adopted, making use of LiDAR, Camera, and Radar sensors. These representations improved 3D object detection performance, but real-world urban scenarios with unexpected situations can still lead to numerous false positives, posing a challenge for robust 3D models. This paper presents a post-processing algorithm that dynamically adjusts object detection thresholds based on the distance from the ego-vehicle. While conventional perception algorithms typically employ a single threshold in post-processing, 3D models perform well in detecting nearby objects but may exhibit suboptimal performance for distant ones. The proposed algorithm tackles this issue by employing adaptive thresholds based on the distance from the ego-vehicle, minimizing false negatives and reducing false positives in the 3D model. The results show performance enhancements in the 3D model across a range of scenarios, encompassing not only typical urban road conditions but also scenarios involving adverse weather conditions.

• 문화기술의 융합
• /
• 제10권1호
• /
• pp.617-623
• /
• 2024

언어와 감정 사이의 복잡한 관계의 특징을 보이며, 우리의 말을 통해 감정을 식별하는 것은 중요한 과제로 인식된다. 이 연구는 음성 및 텍스트 데이터를 모두 포함하는 다중 모드 분류 작업을 통해 음성 언어의 감정을 식별하기 위해 속성 엔지니어링을 사용하여 이러한 과제를 해결하는 것을 목표로 한다. CNN(Convolutional Neural Networks)과 LSTM(Long Short-Term Memory)이라는 두 가지 분류기를 BERT 기반 사전 훈련된 모델과 통합하여 평가하였다. 논문에서 평가는 다양한 실험 설정 전반에 걸쳐 다양한 성능 지표(정확도, F-점수, 정밀도 및 재현율)를 다룬다. 이번 연구 결과는 텍스트와 음성 데이터 모두에서 감정을 정확하게 식별하는 두 모델의 뛰어난 능력을 보인다.

• 산업경영시스템학회지
• /
• 제37권4호
• /
• pp.90-97
• /
• 2014

The mobile SNS is a promising e-service platform of the future. However, measuring its service quality is a challenge. An appropriate model for measuring service quality is required. This paper proposes e-service quality models for analyzing mobile SNS platform quality, based on previous service quality researches. An empirical study is performed on the proposed models. The results show that constructs of existing e-service models such as responsiveness and assurance do not fit the mobile SNS platform, and that loyalty and value are better measures for mobile service quality.

• Parasites, Hosts and Diseases
• /
• 제45권3호
• /
• pp.175-180
• /
• 2007

In order to determine the role of Peyer's patch lymphocytes (PPL) in self-clearing of Cryptosporidium parvum infection in murine models, changes in PPL subsets, their cytokine expression, and in vitro IgG1 and IgA secretions by PPL were observed in primary- and challenge-infected C57BL/6 mice. In primary-infected mice, the percentages of CD4+ T cells, CD8+ T cells, slgA+ B cells, IL-2+ T cells, and $IFN-{\gamma}+$ T cells among the PPL, increased significantly (P < 0.05) on day 10 post-infection (PI). Secretion of IgG1 and IgA in vitro by PPL also increased on day 10 PI. However, all these responses, with the exception of IgG1 and IgA secretions, decreased in challenge-infected mice on day 7 post-challenge (= day 13 PI); their IgG1 and IgA levels were higher (P > 0.05) than those in primary-infected mice. The results suggest that murine PPL play an important role in self-clearing of primary C. parvum infections through proliferation of CD4+, CD8+, IL-2+, and $IFN-{\gamma}+$ T cells, and IgG1 and IgA-secreting 8 cells. In challenge infections, the role of T cells is reduced whereas that of 8 cells secreting IgA appeared to be continuously important.

• Journal of information and communication convergence engineering
• /
• 제21권4호
• /
• pp.346-350
• /
• 2023

Trajectory prediction is an essential element for driving autonomous vehicles, and various trajectory prediction models have emerged with the development of deep learning technology. Convolutional neural network (CNN) is the most commonly used neural network architecture for extracting the features of visual images, and the latest models exhibit high performances. This study was conducted to identify an efficient CNN backbone model among the components of deep learning models for trajectory prediction. We changed the existing CNN backbone network of multiple-trajectory prediction models used as feature extractors to various state-of-the-art CNN models. The experiment was conducted using nuScenes, which is a dataset used for the development of autonomous vehicles. The results of each model were compared using frequently used evaluation metrics for trajectory prediction. Analyzing the impact of the backbone can improve the performance of the trajectory prediction task. Investigating the influence of the backbone on multiple deep learning models can be a future challenge.

• Geomechanics and Engineering
• /
• 재33권6호
• /
• pp.567-581
• /
• 2023

The prediction and achievement of a proper rock fragmentation size is the main challenge of blasting operations in surface mines. This is because an optimum size distribution can optimize the overall mine/plant economics. To this end, this study attempts to develop four improved artificial intelligence models to predict rock fragmentation through cascaded forward neural network (CFNN) and radial basis function neural network (RBFNN) models. In this regards, the CFNN was trained by the Levenberg-Marquardt algorithm (LMA) and Conjugate gradient backpropagation (CGP). Further, the RBFNN was optimized by the Dragonfly Algorithm (DA) and teaching-learning-based optimization (TLBO). For developing the models, the database required was collected from the Midouk copper mine, Iran. After modeling, the statistical functions were computed to check the accuracy of the models, and the root mean square errors (RMSEs) of CFNN-LMA, CFNN-CGP, RBFNN-DA, and RBFNN-TLBO were obtained as 1.0656, 1.9698, 2.2235, and 1.6216, respectively. Accordingly, CFNN-LMA, with the lowest RMSE, was determined as the model with the best prediction results among the four examined in this study.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권11호
• /
• pp.5287-5303
• /
• 2018

Handling vast amount of data found in large transactional datasets is an obvious challenge for the conventional data mining algorithms. Addressing this challenge, our paper proposes a parallel approach for proper decomposition of mining problem into sub-problems in order to find frequent patterns from these datasets. The proposed, Pattern Mining for Large Distributed Dataset (PMLDD) approach, ensures minimum dependencies as well as minimum communications among sub-problems. It establishes a linear aggregation of the intermediate results so that it can be adapted to large-scale programming models like MapReduce. In this context, an algorithmic structure for MapReduce programming model is presented. PMLDD guarantees an efficient load balancing among the sub-problems by a specific selection criterion. Further, it optimizes the number of required iterations over the dataset for mining frequent patterns as compared to the existing approaches. Finally, we believe that our approach is scalable enough to handle larger datasets in terms of performance evaluation, and the result analysis justifies all these mentioned concerns.