• Structural Engineering and Mechanics
• /
• v.63 no.1
• /
• pp.115-124
• /
• 2017

In this paper, a different technique to predict the effects of soil-structure interaction (SSI) on seismic response of building systems is investigated. The technique use a machine learning algorithm called Support Vector Regression (SVR) with technical and analytical results as input features. Normally, the effects of SSI on seismic response of existing building systems can be identified by different types of large data sets. Therefore, predicting and estimating the seismic response of building is a difficult task. It is possible to approximate a real valued function of the seismic response and make accurate investing choices regarding the design of building system and reduce the risk involved, by giving the right experimental and/or numerical data to a machine learning regression, such as SVR. The seismic response of both single-degree-of-freedom system and six-storey RC frame which can be represent of a broad range of existing structures, is estimated using proposed SVR model, while allowing flexibility of the soil-foundation system and SSI effects. The seismic response of both single-degree-of-freedom system and six-storey RC frame which can be represent of a broad range of existing structures, is estimated using proposed SVR model, while allowing flexibility of the soil-foundation system and SSI effects. The results show that the performance of the technique can be predicted by reducing the number of real data input features. Further, performance enhancement was achieved by optimizing the RBF kernel and SVR parameters through grid search.

• Journal of KIISE:Software and Applications
• /
• v.29 no.4
• /
• pp.270-277
• /
• 2002

Text filtering is a task of deciding whether a document has relevance to a specified topic. As Internet and Web becomes wide-spread and the number of documents delivered by e-mail explosively grows the importance of text filtering increases as well. The aim of this paper is to improve the accuracy of text filtering systems by using machine learning techniques. We apply AdaBoost algorithms to the filtering task. An AdaBoost algorithm generates and combines a series of simple hypotheses. Each of the hypotheses decides the relevance of a document to a topic on the basis of whether or not the document includes a certain word. We begin with an existing AdaBoost algorithm which uses weak hypotheses with their output of 1 or -1. Then we extend the algorithm to use weak hypotheses with real-valued outputs which was proposed recently to improve error reduction rates and final filtering performance. Next, we attempt to achieve further improvement in the AdaBoost's performance by first setting weights randomly according to the continuous Poisson distribution, executing AdaBoost, repeating these steps several times, and then combining all the hypotheses learned. This has the effect of mitigating the ovefitting problem which may occur when learning from a small number of data. Experiments have been performed on the real document collections used in TREC-8, a well-established text retrieval contest. This dataset includes Financial Times articles from 1992 to 1994. The experimental results show that AdaBoost with real-valued hypotheses outperforms AdaBoost with binary-valued hypotheses, and that AdaBoost iterated with random weights further improves filtering accuracy. Comparison results of all the participants of the TREC-8 filtering task are also provided.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06c
• /
• pp.168-171
• /
• 2011

One important issue in semantic web is identification and selection of domain concepts for domain ontology learning when several hundreds or even thousands of terms are extracted and available from relevant text documents shared among the members of a domain. We present a novel domain concept acquisition and selection approach for ontology learning that uses affinity propagation algorithm, which takes as input semantic and structural similarity between pairs of extracted terms called data points. Real-valued messages are passed between data points (terms) until high quality set of exemplars (concepts) and cluster iteratively emerges. All exemplars will be considered as domain concepts for learning domain ontologies. Our empirical results show that our approach achieves high precision and recall in selection of domain concepts using less number of iterations.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1996.10a
• /
• pp.253-256
• /
• 1996

Rainfall is one of the major and complicated elements of hydrologic system. Accurate prediction of rainfall is very important to mitigate storm damage. The neural network is a good model to be applied for the classification problem, large combinatorial optimization and nonlinear mapping. In this dissertation, rainfall predictions by the neural network theory were presented. A multi-layer neural network was constructed. The network learned continuous-valued input and output data. The network was used to predict rainfall. The online, multivariate, short term rainfall prediction is possible by means of the developed model. A multidimensional rainfall generation model is applied to Seoul metropolitan area in order to generate the 10-minute rainfall. Application of neural network to the generated rainfall shows good prediction. Also application of neural network to 1-hour real data in Seoul metropolitan area shows slightly good predictions.

• Journal of Information Processing Systems
• /
• v.20 no.3
• /
• pp.360-374
• /
• 2024

A personalized course recommendation algorithm based on deep learning in an online education cloud platform is proposed to address the challenges associated with effective information extraction and insufficient feature extraction. First, the user potential preferences are obtained through the course summary, course review information, user course history, and other data. Second, by embedding, the word vector is turned into a low-dimensional and dense real-valued vector, which is then fed into the compressed interaction network-deep neural network model. Finally, considering that learners and different interactive courses play different roles in the final recommendation and prediction results, an attention mechanism is introduced. The accuracy, recall rate, and F1 value of the proposed method are 0.851, 0.856, and 0.853, respectively, when the length of the recommendation list K is 35. Consequently, the proposed strategy outperforms the comparison model in terms of recommending customized course resources.

• Journal of the Korean Home Economics Association
• /
• v.48 no.9
• /
• pp.113-123
• /
• 2010

This study describes the beginning and further development of a collection web-based materials for an efficient approach to culinary practice education. A database was created using a five-step process of analysis, design, development, operation and evaluation. The menu for the web-based culinary practice educational materials included cooking basics, the real status of cooking, cooking related knowledge, performance evaluation, a data room and a bulletin board. As at 30 July, 2010, the datadase of educational materials, contained a total of 571 items. These comprised 139 cooking pictures, 33 recipes, 22 cooking videos, 74 cooking animations, 57 collections of basic knowledge, 14 evaluation reports, 21 supplementary textbooks, and 211 sets of other related information. The webbased materials are adequate for culinary education purposes, and their use is expected to be very highly valued.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06b
• /
• pp.480-482
• /
• 2012

하이퍼네트워크(Hypernetworks)는 하이퍼에지(hyperedge)들로 이루어진 생성 모델(generative model)로서, 주로 이산(binary) 데이터에 적용되어왔다. 본 논문에서는 이산 데이터와 실수 데이터를 모두 다룰 수 있는 새로운 하이퍼네트워크 모델을 에너지 기반 모델(energy-based model)의 형태로 제시하고, 비지도 학습(unsupervised learning) 알고리즘으로 데이터를 성공적으로 학습함을 간단한 실험을 통해 보이겠다.

• ETRI Journal
• /
• v.37 no.4
• /
• pp.677-684
• /
• 2015

Recently, orthogonal frequency-division multiplexing (OFDM) was applied to VLC systems owing to its high rate capability. On the other hand, a real-valued unipolar OFDM signal for VLC significantly reduces bandwidth efficiency. For practical implementation, channel estimation is required for data demodulation, which causes a further decrease in spectral efficiency. In addition, the large fluctuation of an OFDM signal results in poor illumination quality, such as chromaticity changes. This paper proposes a spectrally efficient method based on a hidden-pilot-aided precoding technology for VLC with less flickering than a conventional OFDM-based method. This approach can obtain channel information without any loss of bandwidth efficiency while ensuring illumination quality by reducing the flickering effect of an OFDM-based VLC. The simulation results show that the proposed method provides a 6.4% gain in bandwidth efficiency with a 4% reduction in flicker compared to a conventional OFDM-based method.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.1
• /
• pp.126-130
• /
• 2005

In this paper, a low-complexity maximum-likelihood (ML) decoder based on QR decomposition, called real-valued LCMLDec decoder or RVLCMLDec for short, is proposed for the Vertical Bell Labs Layered Space-Time (V-BLAST) architecture, a promising candidate for providing high data rates in future fixed wireless communication systems [1]. Computer simulations, in comparison with other detection techniques, show that the proposed decoder is capable of providingthe V-BLAST schemes with ML performance at low detection complexity.

• Journal of information and communication convergence engineering
• /
• v.3 no.2
• /
• pp.76-79
• /
• 2005

In this paper, a low-complexity maximum-likelihood (ML) decoder based on QR decomposition, called real-valued LCMLDec decoder or RVLCMLDec for short, is proposed for the Vertical Bell Labs Layered Space-Time (V-BLAST) architecture, a promising candidate for providing high data rates in future fixed wireless communication systems [1]. Computer simulations, in comparison with other detection techniques, show that the proposed decoder is capable of providing the V­BLAST schemes with ML performance at low detection complexity