• 한국측량학회지
• /
• 제30권6_2호
• /
• pp.653-663
• /
• 2012

Using soft classification, it is possible to obtain the land cover proportions from the remotely sensed image. These land cover proportions are then used as input data for a procedure called "super-resolution mapping" to produce the predicted hard land cover layers at higher resolution than the original remotely sensed image. Superresolution mapping can be implemented using a number of algorithms in which the Hopfield Neural Network (HNN) has showed some advantages. The HNN has improved the land cover classification through superresolution mapping greatly with the high resolution data. However, the super-resolution mapping is based on the spatial dependence assumption, therefore it is predicted that the accuracy of resulted land cover classes depends on the relative size of spatial features and the spatial resolution of the remotely sensed image. This research is to evaluate the capability of HNN to implement the super-resolution mapping for SPOT image to create higher resolution land cover classes with different zoom factor.

• 한국정보처리학회논문지
• /
• 제6권5호
• /
• pp.1296-1302
• /
• 1999

In this paper, we propose an algorithms for the mounted PCB classification system using wavelet transform and ART2 neural network. The feature informations of a mounted PCB can be extracted from the coefficient matrix of wavelet transform adapted subband concept. As the preprocessing process, only the PCB area in the input image is extracted by histogram method and the feature vectors are composed of using wavelet transform method. These feature vectors are used as the input vector of ART2 neural network. In the experiment using 55 mounted PCB images, the proposed algorithm shows 100% classification rate at the vigilance parameter $\rho$=0.99. The proposed algorithm has some advantages of the feature extraction in the compressed domain and the simplification of processing steps.

• 대한설비공학회:학술대회논문집
• /
• 대한설비공학회 2006년도 하계학술발표대회 논문집
• /
• pp.454-459
• /
• 2006

There are two main streams for pattern classification technology One is the method using PCA (Principal Component Analysis) and the other is the method using Neural network. Both of them have merits and demerits. In general, using PCA is so simple while using neural network can improve algorithm continually. Algorithm using neural network needs so many calculations rendering very slow response. In this work, an attempt is made to develop algorithms adopting both PCA and neural network merits for simpler, but faster and smarter.

• 센서학회지
• /
• 제30권5호
• /
• pp.286-289
• /
• 2021

Image classification is one of the fundamental applications of computer vision. It enables a system to identify an object in an image. Recently, image classification applications have broadened their scope from computer applications to edge devices. The convolutional neural network (CNN) is the main class of deep learning neural networks that are widely used in computer tasks, and it delivers high accuracy. However, CNN algorithms use a large number of parameters and incur high computational costs, which hinder their implementation in edge hardware devices. To address this issue, this paper proposes a lightweight image classifier that provides good accuracy while using fewer parameters. The proposed image classifier diverts the input into three paths and utilizes different scales of receptive fields to extract more feature maps while using fewer parameters at the time of training. This results in the development of a model of small size. This model is tested on the CIFAR-10 dataset and achieves an accuracy of 90% using .26M parameters. This is better than the state-of-the-art models, and it can be implemented on edge devices.

• 한국멀티미디어학회논문지
• /
• 제21권12호
• /
• pp.1387-1395
• /
• 2018

Core algorithm of deep learning Convolutional Neural Network(CNN) shows better performance than other machine learning algorithms. However, if there is not sufficient data, CNN can not achieve satisfactory performance even if the classifier is excellent. In this situation, it has been proven that the use of transfer learning can have a great effect. In this paper, we apply two transition learning methods(freezing, retraining) to three CNN models(ResNet-50, Inception-V3, DenseNet-121) and compare and analyze how the classification performance of CNN changes according to the methods. As a result of statistical significance test using various evaluation indicators, ResNet-50, Inception-V3, and DenseNet-121 differed by 1.18 times, 1.09 times, and 1.17 times, respectively. Based on this, we concluded that the retraining method may be more effective than the freezing method in case of transition learning in image classification problem.

• 대한상한금궤의학회지
• /
• 제14권1호
• /
• pp.41-50
• /
• 2022

Objective : The purpose of this study is to explore the most suitable machine learning model algorithm for Shanghanlun diagnostic system classification using natural language processing (NLP). Methods : A total of 201 data items were collected from 『Shanghanlun』 and 『Clinical Shanghanlun』, 'Taeyangbyeong-gyeolhyung' and 'Eumyangyeokchahunobokbyeong' were excluded to prevent oversampling or undersampling. Data were pretreated using a twitter Korean tokenizer and trained by logistic regression, ridge regression, lasso regression, naive bayes classifier, decision tree, and random forest algorithms. The accuracy of the models were compared. Results : As a result of machine learning, ridge regression and naive Bayes classifier showed an accuracy of 0.843, logistic regression and random forest showed an accuracy of 0.804, and decision tree showed an accuracy of 0.745, while lasso regression showed an accuracy of 0.608. Conclusions : Ridge regression and naive Bayes classifier are suitable NLP machine learning models for the Shanghanlun diagnostic system classification.

• 한국정보과학회논문지:정보통신
• /
• 제37권3호
• /
• pp.175-186
• /
• 2010

다양한 어플리케이션의 등장과 인터넷 사용자의 급속한 성장으로 인하여, 인터넷 라우터는 패킷이 입력되는 속도와 같은 속도로 패킷 분류작업을 수행하여 패킷의 클래스에 따른 품질 보장을 제공할 것이 요구되고 있다. 패킷 분류란 라우터에 입력된 패킷의 헤더가 가지고 있는 여러 개의 필드에 대해 다차원 검색을 수행하여, 미리 정의된 룰과 일치하는 결과 가운데 최우선순위를 갖는 룰을 찾아내는 과정을 말한다. 빠른 패킷 분류를 위하여 다양한 패킷 분류 알고리즘이 제안되어오고 있으며, 튜플 공간 제거(tuple space pruning) 알고리즘은 일치 가능한 룰을 갖는 튜플들만을 해싱을 사용하여 검색함으로 빠른 검색 성능을 제공한다. 블룸 필터(Bloom filter)는 특정 집합에 속하는 원소들의 멤버쉽에 관한 정보를 간단한 비트-벡터로 표현하는 데이터 구조로서, 특정 입력 값이 집합에 속한 원소인지를 알려주는 선-필터(pre-filter)로 사용된다. 본 논문에서는 블룸 필터를 이용하여 일치 가능성이 없는 튜플을 효율적으로 제거하는 새로운 튜플 제거 알고리즘을 제안한다. 실제 라우터에서 사용되는 룰 셋과 비슷한 특성을 갖는다고 알려진 데이터 베이스에 대한 성능 비교를 통하여, 본 논문에서 제안하는 구조가 패킷 분류 성능 및 메모리 사용량에 있어서 기존의 튜플공간 제거 알고리즘과 비교하여 월등히 우수함을 보았다.

• 터널과지하공간
• /
• 제31권6호
• /
• pp.494-507
• /
• 2021

TBM의 활용이 증가하면서 최근 국내에서도 머신러닝 기법으로 TBM 데이터를 분석하여 TBM 전방의 지반을 예측하고 디스크커터의 교환주기 예측 및 굴진율을 예측하는 연구가 수행되고 있다. 본 연구에서는 TBM 굴진 시 기계 데이터를 대상으로 전통적 암반에 대한 분류 기법과 최근에 다양한 분야에서 널리 사용되고 있는 머신러닝 기법들을 접목하여 슬러리 쉴드 TBM 현장의 암반 특성에 대한 분류 예측을 하였다. 암반 특성 분류 기준 항목을 RQD, 일축압축강도, 탄성파속도로 설정하고 항목별 암반상태를 클래스 0(양호),1(보통),2(불량)의 3개 클래스로 구분한 다음, 6개의 분류 알고리즘에 대한 기계학습을 수행하였다. 그 결과, 앙상블 계열의 모델이 좋은 성능을 보여주었고 특히 학습성능과 더불어 학습속도에서 우수한 결과를 보인 LigthtGBM 모델이 대상 현장 지반에서 최적인 것으로 나타났다. 본 연구에서 설정한 3가지 암반 특성에 대한 분류 모델을 활용하면 지반정보가 제공되지 않은 구간에 대한 암반 상태를 제공할 수 있어 굴착작업 시 도움을 줄 수 있을 것으로 판단된다.

• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 1999년도 추계학술대회-지능형 정보기술과 미래조직 Information Technology and Future Organization
• /
• pp.365-373
• /
• 1999

Recently, numerous studies have demonstrated that artificial intelligence such as neural networks can be an alternative methodology for classification problems to which traditional statistical methods have long been applied. In building neural network model, the selection of independent and dependent variables should be approached with great care and should be treated as a model construction process. Irrespective of the efficiency of a learning procedure in terms of convergence, generalization and stability, the ultimate performance of the estimator will depend on the relevance of the selected input variables and the quality of the data used. Approaches developed in statistical methods such as correlation analysis and stepwise selection method are often very useful. These methods, however, may not be the optimal ones for the development of neural network models. In this paper, we propose a genetic algorithms approach to find an optimal or near optimal input variables for neural network modeling. The proposed approach is demonstrated by applications to bankruptcy prediction modeling. Our experimental results show that this approach increases overall classification accuracy rate significantly.

• 제어로봇시스템학회논문지
• /
• 제3권2호
• /
• pp.179-186
• /
• 1997

In this paper we suggest a development of surface defect inspection algorithms for cold mill strip. The defects which exist in a surface of cold mill strip have a scattering or singular distribution. This paper consists of preprocessing, feature extraction and defect classification. By preprocessing, the binarized defect image is achieved. In this procedure, Top-hit transform, adaptive thresholding, thinning and noise rejection are used. Especially, Top-hit transform using local min/max operation diminishes the effect of bad lighting. In feature extraction, geometric, moment and co-occurrence matrix features are calculated. For the defect classification, multilayer neural network is used. The proposed algorithm showed 15% error rate.