• 한국컴퓨터정보학회논문지
• /
• 제27권5호
• /
• pp.77-84
• /
• 2022

본 논문에서는 새로운 비지도 특징 선별 기법을 제안한다. 기존 비지도 방식의 특징 선별 기법들은 특징을 선별하기 위해 가상의 레이블 데이터를 정하고 주어진 데이터를 이 레이블 데이터에 사영하는 회귀 분석 방식으로 특징을 선별하였다. 하지만 가상의 레이블은 데이터로부터 생성되기 때문에 사영된 공간이 비슷하게 형성될 수 있다. 따라서 기존의 방법들에서는 제한된 공간에서만 특징이 선택될 수 있었다. 이를 해소하기 위해 본 논문에서는 직교 사영과 저랭크 근사를 이용하여 특징을 선별한다. 이 문제를 해소하기 위해 가상의 레이블을 직교 사영하고 이 공간에 데이터를 사영할 수 있도록 한다. 이를 통해 더 주요한 특징 선별을 기대할 수 있다. 그리고 사영을 위한 변환 행렬에 저랭크 제한을 두어 더 효과적으로 저차원 공간의 특징을 선별할 수 있도록 한다. 이 목표를 달성하기 위해 본 논문에서는 비용 함수를 설계하고 효율적인 최적화 방법을 제안한다. 여섯 개의 데이터에 대한 실험 결과는 제안된 방법이 대부분의 경우 기존의 비지도 특징 선별 기법보다 좋은 성능을 보여주었다.

• 대한산업공학회지
• /
• 제40권3호
• /
• pp.275-282
• /
• 2014

One of the most widely used methods for dimensionality reduction is principal component analysis (PCA). However, the reduced dimensions from PCA do not provide a clear interpretation with respect to the original features because they are linear combinations of a large number of original features. This interpretation problem can be overcome by feature selection approaches that identifying the best subset of given features. In this study, we propose an unsupervised feature selection method based on the geometrical information of PCA loading vectors. Experimental results from a simulation study demonstrated the efficiency and usefulness of the proposed method.

• 정보처리학회논문지B
• /
• 제8B권6호
• /
• pp.675-684
• /
• 2001

본 논문은 태그가 없는 단순 말뭉치만을 가지고 자율학습을 이용하여 정보 검색을 위한 색인어의 추출 등에 이용될 수 있도록 한국어의 실질 형태소와 형식 형태소를 분리해내는 기법에 대하여 기술한다. 본 기법은 사전 등의 언어 관련 지식을 요구하지 않으며 오직 단순 말뭉치만을 필요로 한다. 또한 자율학습을 이용함으로써 사람의 간섭이 필요하지 않아 학습에 필요한 시간과 노력이 거의 들지 않는다. 본 방식은 잘 확립된 통계적 방법론을 이용하기 때문에 일반적인 휴리스틱과는 달리 이론적인 기반이 확고하여 확장 및 발전이 용이하다. 본 결과는 한국어에 우선 적용되었으나 한국어에 종속적인 방법이 아니어서 다른 교착어에도 쉽게 적용될 수 있을 것이다.

• Nuclear Engineering and Technology
• /
• 제55권2호
• /
• pp.506-515
• /
• 2023

This study is to find a method to learn engine sound after the start-up of a diesel generator installed in nuclear power plant with an unsupervised deep learning algorithm (CNN autoencoder) and a new method to predict the failure of a diesel generator using it. In order to learn the sound of a diesel generator with a deep learning algorithm, sound data recorded before and after the start-up of two diesel generators was used. The sound data of 20 min and 2 h were cut into 7 s, and the split sound was converted into a spectrogram image. 1200 and 7200 spectrogram images were created from sound data of 20 min and 2 h, respectively. Using two different deep learning algorithms (CNN autoencoder and binary classification), it was investigated whether the diesel generator post-start sounds were learned as normal. It was possible to accurately determine the post-start sounds as normal and the pre-start sounds as abnormal. It was also confirmed that the deep learning algorithm could detect the virtual abnormal sounds created by mixing the unusual sounds with the post-start sounds. This study showed that the unsupervised anomaly detection algorithm has a good accuracy increased about 3% with comparing to the binary classification algorithm.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2004년도 춘계학술발표회논문집
• /
• pp.159-164
• /
• 2004

SAM(Spectral Angle Mapper) is the method using the similarly of the angle between pairs of signatures instead of the spectral distance(MDC, MLC etc.) for classification or clustering. In this paper, we applied unsupervised techniques(Unsupervised SAM and ISODATA) to the Hyperspectral Image(Hyperion) which has innumerable, narrow and contiguous spectral bands and Multispectral Image(ETM$\^$+/) for the clustering of signatures. The overall measured accuracies of the USAM and ISODATA of multispectral image were 76.52%, 53.91% and the USAM and ISODATA of hyperspectral image were 63.04%, 53.91%. From the results of our test, we report that the Unsupervised SAM is better classfication technique than ISODATA. Also we believe that the "Spectral Angle" can potentially be one of the most accurate classifier not only multispectral images but hyperspectral images.

• Smart Structures and Systems
• /
• 제30권6호
• /
• pp.557-569
• /
• 2022

Vibration-based structural health monitoring (SHM) is crucial for the dynamic maintenance of civil building structures to protect property security and the lives of the public. Analyzing these vibrations with modern artificial intelligence and deep learning (DL) methods is a new trend. This paper proposed an unsupervised deep learning method based on a convolutional autoencoder (CAE), which can overcome the limitations of conventional supervised deep learning. With the convolutional core applied to the DL network, the method can extract features self-adaptively and efficiently. The effectiveness of the method in detecting damage is then tested using a benchmark model. Thereafter, this method is used to detect damage and instant disaster events in a rubber bearing-isolated gymnasium structure. The results indicate that the method enables the CAE network to learn the intact vibrations, so as to distinguish between different damage states of the benchmark model, and the outcome meets the high-dimensional data distribution characteristics visualized by the t-SNE method. Besides, the CAE-based network trained with daily vibrations of the isolating layer in the gymnasium can precisely recover newly collected vibration and detect the occurrence of the ground motion. The proposed method is effective at identifying nonlinear variations in the dynamic responses and has the potential to be used for structural condition assessment and safety warning.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2010년도 추계학술발표대회
• /
• pp.625-628
• /
• 2010

This paper proposes a novel feature extraction method for unsupervised multispectral image segmentation based in one dimensional combined neighborhood differences (1D CND). In contrast with the original CND, which is applied with traditional image, 1D CND is computed on a single pixel with various bands. The proposed algorithm utilizes the sign of differences between bands of the pixel. The difference values are thresholded to form a binary codeword. A binomial factor is assigned to these codeword to form another unique value. These values are then grouped to construct the 1D CND feature image where is used in the unsupervised image segmentation. Various experiments using two LANDSAT multispectral images have been performed to evaluate the segmentation and classification accuracy of the proposed method. The result shows that 1D CND feature outperforms the spectral feature, with average classification accuracy of 87.55% whereas that of spectral feature is 55.81%.

• Journal of Computing Science and Engineering
• /
• 제11권4호
• /
• pp.121-129
• /
• 2017

Feature selection has been widely established as an efficient technique for microarray data analysis. Feature selection aims to search for the most important feature/gene subset of a given dataset according to its relevance to the current target. Unsupervised feature selection is considered to be challenging due to the lack of label information. In this paper, we propose a novel method for unsupervised feature selection, which incorporates embedded learning and $l_{2,1}-norm$ sparse regression into a framework to select genes in microarray data analysis. Local tangent space alignment is applied during embedded learning to preserve the local data structure. The $l_{2,1}-norm$ sparse regression acts as a constraint to aid in learning the gene weights correlatively, by which the proposed method optimizes for selecting the informative genes which better capture the interesting natural classes of samples. We provide an effective algorithm to solve the optimization problem in our method. Finally, to validate the efficacy of the proposed method, we evaluate the proposed method on real microarray gene expression datasets. The experimental results demonstrate that the proposed method obtains quite promising performance.

• 인지과학
• /
• 제10권1호
• /
• pp.13-24
• /
• 1999

논문에서는 모듈화 신경 을 이용한 비지도 학습방법의 분류기를 제안한다. 각 모듈은 데이터의 통계학적인 분석의 결과로 설계되어져서, 데이터의 독립적인 군집들을 나타내게 된다. 이런 신경의 독립적인 분류 결과와 근접거리 척도를 이용한 유사도 측정을 통해 더욱 정확한 분류를 가능케 하며, 오 분류를 하는 모듈을 삭제함으로써 계산 을 줄인다. 이런 과정을 통해 신경 에 사용되는 각종 변수에 대한 별다른 조사 과정 없이 최상의 성능을 발휘하는 신경 에 준 는 성능을 가진 신경 망을 구축했다.

• 전기전자학회논문지
• /
• 제24권4호
• /
• pp.1011-1016
• /
• 2020

We propose a framework called Stacked Gated Recurrent Unit - Infrequent Residual Analysis (SG-IRA) that detects anomalies in time-series data that can be trained on streams of raw sensor data without any pre-labeled dataset. To enable such unsupervised learning, SG-IRA includes an estimation model that uses a stacked Gated Recurrent Unit (GRU) structure and an analysis method that detects anomalies based on the difference between the estimated value and the actual measurement (residual). SG-IRA's residual analysis method dynamically adapts the detection threshold from the population using frequency analysis, unlike the baseline model that relies on a constant threshold. In this paper, SG-IRA is evaluated using the industrial control systems (ICS) datasets. SG-IRA improves the detection performance (F1 score) by 5.9% compared to the baseline model.