• Proceedings of the Korea Water Resources Association Conference
• /
• 2008.05a
• /
• pp.1727-1731
• /
• 2008

수자원의 효율적인 관리를 위해서는 신뢰성 있는 유량자료의 획득이 대단히 중요하다. 우리나라는 양질의 유량자료를 획득하기 위해 매년 많은 시간과 돈을 투자하고 있으나 자료의 질적인 면에서 만족할 만한 성과를 얻지 못하고 있다. 현재까지 우리나라의 유량자료는 댐의 수문자료와 수량관리 부처인 건교부에서 운영하는 수위표 지점의 수위-유량곡선에서 산출된 자료에 의존하고 있다. 그러나 수위-유량 관계식을 보정하기 위한 유량측정사업이 지속적이지 못하며, 이 관계식은 유량이 적은 저수기 및 갈수기에는 부정확하다는 한계가 있다. 또한, 국립환경과학원 낙동강물환경연구소에서 오염총량관리를 위한 낙동강수계 유량측정사업을 실시하고 있지만, 목적은 낙동강수계의 오염총량관리 단위유역 말단 47개 지점에서 유량측정을 효율적으로 실시하여 수질정책의 기초자료를 제공하는데 있다. 이 자료 역시 오염총량관리를 위하여 유량측정을 실시하여 수자원의 효율적인 관리를 위한 일 유량을 알 수가 없는 한계점을 가지고 있다. 따라서 저수기 및 갈수기에 수질정책의 기초자료를 제공하기 위해서 하천을 포함한 유역의 정확한 강우-유출특성의 파악이 필요하다. 그러나 강우-유출특성 또한 유역 내 강우의 시 공간적 분포가 다르며 그 자가 비선형성이 강하고 여러 변동성을 포함하므로, 강우로부터 하천의 유출량의 정확한 해석이 불가능하다. 그러나 최근 인공지능 분야에서 신호처리, 지능제어 및 패턴인식 등의 수단으로 사용되고 있는 신경망은 학습이라는 최적화 과정을 통해 입력과 출력으로 구성되는 하나의 시스템을 비선형적으로 구축할 수 있으며 이러한 이점을 활용하여 수자원 분야에서 다양하게 적용되고 있다. 본 연구의 목적은 강우-유출자료 및 댐 방류량 자료의 비선형적인 특정을 가장 잘 반영할 수 있는 신경망모형을 적용하여 수질정책의 기초자료를 제공하기 위하여 신뢰성 있는 유량자료를 산정하는 모형을 개발하는 것이다. 이를 위해서 낙동강물환경연구소에서 오염총량관리를 위한 낙동강수계 유량측정 지점 상류의 댐 방류량의 일 방류량자료와 강우자료를 입력 자료로 하여 유량을 예측할 수 있는 유량예측 신경망 모형 FFBN(Flow Forecasting By Neural)을 개발하였다. 그리고 입력 자료로서 장기유출모형인 SWAT의 모의결과를 입력 자료로 추가한 FFBNS(Flow Forecasting By Neural and SWAT)을 개발하였다. 신경망 모형의 구조는 입력층과 출력층 사이에 하나의 은닉층이 존재하는 다층 신경망으로 구성하였으며, 학습단계에서는 오류 역전파 알고리듬 학습방법 중 모멘텀법을 사용하였다. 예측된 유출량을 실측치와의 비교를 위하여 낙본D지점과 낙본 E지점에 대하여 $2005{\sim}2006$년까지의 모의 결과를 낙동 수위측정지점과 구미 수위측정지점의 실측치 통하여 복잡한 비선형성을 가지는 유출 시계열 자료에 대한 효과적인 최적의 신경망모델을 개발하여 유량을 예측하고 적용 가능성을 검토하고자 한다. 모의 결과는 수질정책의 기초자료 제공에 기여할 수 있을 것으로 판단된다.

• Journal of KIISE:Software and Applications
• /
• v.33 no.11
• /
• pp.986-998
• /
• 2006

There has been an increasing necessity of Semantic Web-based metadata that helps computers efficiently understand and manage an information increased with the growth of Internet. However, it seems inevitable to face some semantically ambiguous information when metadata is generated. Therefore, we need a solution to this problem. This paper proposes a new method for automated metadata generation with the help of a concept of class, in which some ambiguous words imbedded in information such as documents are semantically more related to others, by using probability model of consequent words. We considers ambiguities among defined concepts in ontology and uses the Hidden Markov Model to be aware of part of a named entity. First of all, we constrict a Markov Models a better understanding of the named entity of each class defined in ontology. Next, we generate the appropriate context from a text to understand the meaning of a semantically ambiguous word and solve the problem of ambiguities during generating metadata by searching the optimized the Markov Model corresponding to the sequence of words included in the context. We experiment with seven semantically ambiguous words that are extracted from computer science thesis. The experimental result demonstrates successful performance, the accuracy improved by about 18%, compared with SemTag, which has been known as an effective application for assigning a specific meaning to an ambiguous word based on its context.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.9
• /
• pp.75-84
• /
• 2012

HMM-based Text-to-Speech systems generally utilize context dependent tri-phone units from a large corpus speech DB to enhance the synthetic speech. To downsize a large corpus speech DB, acoustically similar tri-phone units are clustered based on the decision tree using context dependent information. Context dependent information includes phoneme sequence as well as prosodic information because the naturalness of synthetic speech highly depends on the prosody such as pause, intonation pattern, and segmental duration. However, if the prosodic information was complicated, many context dependent phonemes would have no examples in the training data, and clustering would provide a smoothed feature which will generate unnatural synthetic speech. In this paper, instead of complicate prosodic information we propose a simple three prosodic boundary types and decision tree questions that use rising tone, falling tone, and monotonic tone to improve naturalness. Experimental results show that our proposed method can improve naturalness of a HMM-based Korean TTS and get high MOS in the perception test.

• The Korean Journal of Applied Statistics
• /
• v.34 no.5
• /
• pp.745-760
• /
• 2021

Prediction of a stock price has been a subject of interest for a long time in financial markets, and thus, many studies have been conducted in various directions. As the efficient market hypothesis introduced in the 1970s acquired supports, it came to be the majority opinion that it was impossible to predict stock prices. However, recent advances in predictive models have led to new attempts to predict the future prices. Here, we summarize past studies on the price prediction by evaluation measures, and predict the direction of stock prices of Samsung Electronics, LG Chem, and NAVER by applying various machine learning models. In addition to widely used technical indicator variables, accounting indicators such as Price Earning Ratio and Price Book-value Ratio and outputs of the hidden Markov Model are used as predictors. From the results of our analysis, we conclude that no models show significantly better accuracy and it is not possible to predict the direction of stock prices with models used. Considering that the models with extra predictors show relatively high test accuracy, we may expect the possibility of a meaningful improvement in prediction accuracy if proper variables that reflect the opinions and sentiments of investors would be utilized.

• Journal of the Korea Computer Graphics Society
• /
• v.8 no.3
• /
• pp.17-23
• /
• 2002

Motion capture system is widely used nowadays in the entertainment industry like movies, computer games and broadcasting. This system consist of several high resolution and high speed CCD cameras and expensive frame grabbing hardware for image acquisition. KAIST VR laboratory focused on low cost system for a few years and have been developed a LAN based optical motion capture system. But, by using low cost system some problems like occlusion, noise and swapping of markers' trajectory can be occurred. And more labor intensive work is needed for post-processing process. In this thesis, we propose a trajectory rectification algorithm by confidence model of markers attached on actor. Confidence model is based on graph structure and consist of linkage, marker and frame confidence. To reduce the manual work in post-processing, we have to reconstruct the marker graph by maximizing the frame confidence.

• Journal of KIISE:Software and Applications
• /
• v.30 no.9
• /
• pp.867-875
• /
• 2003

This paper concerns continuous density HMM topology optimization. There have been several researches for HMM topology optimization. BIC (Bayesian Information Criterion) is one of the well known optimization criteria, which assumes statistically well behaved homogeneous model parameters. HMMs, however, are composed of several different kind of parameters to accommodate complex topology, thus BIC's assumption does not hold true for HMMs. Even though BIC reduced the total number of parameters of HMMs, it could not improve the recognition rates. In this paper, we proposed two new model selection criteria, HBIC (HMM-oriented BIC) and BIC_Anti. The former is proposed to improve BIC by estimating model priors separately. The latter is to combine BIC and anti-likelihood to accelerate discrimination power of HMMs. We performed some comparative research on couple of model selection criteria for online handwriting data recognition. We got better recognition results with fewer number of parameters.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.2 s.314
• /
• pp.82-92
• /
• 2007

This paper describes a novel global shape (GS) feature based on radial basis function network (RBFN) and the extraction method of the proposed feature for 3D face recognition. RBFN is the weighted sum of RBfs, it well present the non-linearity of a facial shape using the linear combination of RBFs. It is the proposed facial feature that the weights of RBFN learned by the horizontal profiles of a face. RBFN based feature expresses the locality of the facial shape even if it is GS feature, and it reduces the feature complexity like existing global methods. And it also get the smoothing effect of the facial shape. Through the experiments, we get 94.7% using the proposed feature and hidden markov model (HMM) to match the features for 100 gallery set with those for 300 test set.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.12 no.6
• /
• pp.511-516
• /
• 2002

The neural networks may have problem such that the amount of calculation for the network learning goes too big according to the dimension of the dimension. To overcome this problem, the wavelet neural networks(WNN) which use the orthogonal basis function in the hidden node are proposed. One can compose wavelet functions as activation functions in the WNN by determining the scale and center of wavelet function. In this paper, when we compose the WNN using wavelet functions, we set a single scale function as a node function together. We intend that one scale function approximates the target function roughly, the other wavelet functions approximate it finely During the determination of the parameters, the wavelet functions can be determined by the global search for solutions suitable for the suggested problem using the genetic algorithm and finally, we use the back-propagation algorithm in the learning of the weights.

• KIISE Transactions on Computing Practices
• /
• v.23 no.4
• /
• pp.244-249
• /
• 2017

Steganalysis is to detect information hidden by steganography inside general data such as images. There are stegoanalysis techniques that use machine learning (ML). Existing ML approaches to steganalysis are based on extracting features from stego images and modeling them. Recently deep learning-based methodologies have shown significant improvements in detection accuracy. However, all the existing methods, including deep learning-based ones, have a critical limitation in that they can only detect stego images that are created by a specific steganography method. In this paper, we propose a generalized steganalysis method that can model multiple types of stego images using deep learning. Through various experiments, we confirm the effectiveness of our approach and envision directions for future research. In particular, we show that our method can detect each type of steganography with the same level of accuracy as that of a steganalysis method dedicated to that type of steganography, thereby demonstrating the general applicability of our approach to multiple types of stego images.

• Journal of KIISE:Information Networking
• /
• v.30 no.2
• /
• pp.207-221
• /
• 2003

A hash chain is highly efficient, attractive structure to use in electronic cash. Previous systems using hash chain were, however, either credit-based vendor-specific cash or debit-based general-purpose cash which lacks efficiency due to double spending problem. In this paper, we propose a new divisible cash system using hash chain. This newly proposed cash is general-purpose, debit-based, anonymous, and offline. The efficiency of the system results from its capacity to pay variable amounts with no additional costs. A client always performs a single blind signature in the withdrawal phase, independent of the length of the chain. During payment, a client performs a single challenge-and-response or generates a single signature, independent of the amount paid. This system provides a new refund mechanism, which uses a refund ticket, that allows clients to refund the unspent part of the chain without revealing any connection to the spent part.