• 제어로봇시스템학회논문지
• /
• 제16권12호
• /
• pp.1150-1158
• /
• 2010

In this study, the Polynomial-based Radial Basis Function Neural Networks is proposed as one of the recognition part of overall face recognition system that consists of two parts such as the preprocessing part and recognition part. The design methodology and procedure of the proposed pRBFNNs are presented to obtain the solution to high-dimensional pattern recognition problem. First, in preprocessing part, we use a CCD camera to obtain a picture frame in real-time. By using histogram equalization method, we can partially enhance the distorted image influenced by natural as well as artificial illumination. We use an AdaBoost algorithm proposed by Viola and Jones, which is exploited for the detection of facial image area between face and non-facial image area. As the feature extraction algorithm, PCA method is used. In this study, the PCA method, which is a feature extraction algorithm, is used to carry out the dimension reduction of facial image area formed by high-dimensional information. Secondly, we use pRBFNNs to identify the ID by recognizing unique pattern of each person. The proposed pRBFNNs architecture consists of three functional modules such as the condition part, the conclusion part, and the inference part as fuzzy rules formed in 'If-then' format. In the condition part of fuzzy rules, input space is partitioned with Fuzzy C-Means clustering. In the conclusion part of rules, the connection weight of pRBFNNs is represented as three kinds of polynomials such as constant, linear, and quadratic. Coefficients of connection weight identified with back-propagation using gradient descent method. The output of pRBFNNs model is obtained by fuzzy inference method in the inference part of fuzzy rules. The essential design parameters (including learning rate, momentum coefficient and fuzzification coefficient) of the networks are optimized by means of the Particle Swarm Optimization. The proposed pRBFNNs are applied to real-time face recognition system and then demonstrated from the viewpoint of output performance and recognition rate.

• 한국멀티미디어학회논문지
• /
• 제23권11호
• /
• pp.1349-1360
• /
• 2020

In recent years, Convolutional Neural Networks (CNNs) have been successfully implemented in different tasks of computer vision. Since CNN models are the representatives of supervised learning algorithms, they demand large amount of data in order to train the classifiers. Thus, obtaining data with correct labels is imperative to attain the state-of-the-art performance of the CNN models. However, labelling datasets is quite tedious and expensive process, therefore real-life datasets often exhibit incorrect labels. Although the issue of poorly labelled datasets has been studied before, we have noticed that the methods are very complex and hard to reproduce. Therefore, in this research work, we propose Deep CleanNet - a considerably simple system that achieves competitive results when compared to the existing methods. We use K-means clustering algorithm for selecting data with correct labels and train the new dataset using a deep CNN model. The technique achieves competitive results in both training and validation stages. We conducted experiments using MNIST database of handwritten digits with 50% corrupted labels and achieved up to 10 and 20% increase in training and validation sets accuracy scores, respectively.

• 전기학회논문지
• /
• 제62권5호
• /
• pp.696-704
• /
• 2013

In this paper, we introduce the LED emotional lighting system realized with the aid of both intelligent algorithm and RGB LED combined with White LED. Generally, the illumination is known as a design factor to form the living place that affects human's emotion and action in the light- space as well as the purpose to light up the specific space. The LED emotional lighting system that can express emotional atmosphere as well as control the quantity of light is designed by using both RGB LED to form the emotional mood and W LED to get sufficient amount of light. RBFNNs is used as the intelligent algorithm and the network model designed with the aid of LED control parameters (viz. color coordinates (x and y) related to color temperature, and lux as inputs, RGBW current as output) plays an important role to build up the LED emotional lighting system for obtaining appropriate color space. Unlike conventional RBFNNs, Fuzzy C-Means(FCM) clustering method is used to obtain the fitness values of the receptive function, and the connection weights of the consequence part of networks are expressed by polynomial functions. Also, the parameters of RBFNN model are optimized by using PSO(Particle Swarm Optimization). The proposed LED emotional lighting can save the energy by using the LED light source and improve the ability to work as well as to learn by making an adequate mood under diverse surrounding conditions.

• 한국지능시스템학회논문지
• /
• 제9권4호
• /
• pp.444-450
• /
• 1999

본 논문에서는 칼라로 입력된 문서 영상에서 문자 영역추출을 위하여 k-means을 이용한 클러스트링 알고리즘을 제안하였다. 칼라 영상의 클러스트링을 위해서 HIS 좌표계에 적합한 거리함수를 제안하였다. 이를 인식하기 위한 전처리 단계인 문자분리(segmentation)방법은 연결 화소를 이용한 개별문자 추출 알고리즘을 제안하였다. 본 알고리즘 에서는 문자분리방벙에서 접촉문자 (touching character)또는 겹친 문자(overlapped character)등과 같이 분리가 곤란한 문자를 개별문자로 분리하는 방법이다. 기존의 문자 분리방법에서는 투영(projection)dop 의한 방법과 외곽선(edge)추적에 의한 방법등을 사용하여 왔으나 제안된 방법은 문자열 추출후 한번의 투영으로 연결화소를 이용하여 개별문자를 추출한다. 문자 영역과 비 문자 영역을 구분하여 개발문자 추출을 한 결과 단순한 이진 영상이 아닌 칼라 영상에서의 문서 처리가 큰 의의가 있고 기존의 문서 처리기 보다 향상된 알고리즘인 것을 확인하였다.

• 지능정보연구
• /
• 제20권4호
• /
• pp.107-120
• /
• 2014

본 연구에서는 사회연결망분석기법 중 하나인 구조적 공백 분석 결과를 이용하여 추천과정에 사용자의 정성적이고 감성적인 정보를 반영할 수 있는 협업필터링 기반의 추천시스템을 제안한다. 협업필터링은 추천기술 중 가장 많이 활용되고 있지만 전통적으로 확장성과 희박성 등의 문제점뿐 만 아니라 사용자-상품 매트릭스의 선호도만을 이용하여 추천을 함으로써 사용자의 정성적이고 감성적인 정보를 추천과정에 반영하지 못한다는 한계점이 있다. 본 연구에서 제안하는 추천시스템은 사회연결망분석에서 중심성 분석과 함께 연결망 내의 주요개체를 탐지할 수 있는 구조적 공백 분석을 이용하여 연결망 내의 대표 사용자들을 추출한 후 이들을 중심으로 군집을 형성한 후 각 군집색인 협업필터링을 수행하는 과정을 통해 전통적인 협업필터링에서 반영하지 못했던 정성적, 감성적 정보를 반영한다. 한편, 군집색인 협업필터링을 수행함으로써 추천의 효율성을 높일 수 있는 장점도 있다. 본 연구에서는 실제 사용자들의 상품에 대한 선호도 평가점수와 사용자들의 사회연결망 정보를 수집하여 실험을 수행하고 전통적인 협업필터링과 다양한 형태의 협업필터링과의 추천성과 비교를 통하여 제안하는 시스템의 유용성을 확인한다. 비교모형으로는 전통적인 협업필터링, 임의 군집색인 기반 협업필터링, k평균 군집색인 기반 협업필터링을 이용한 추천시스템이며, 실험 결과, 제안한 모형이 다른 비교모형에 비해 추천성과의 정확도가 가장 우수하였다. 추천성과의 차이에 대한 통계적 유의성 검정 결과, 제안 모형은 전통적인 협업필터링 기반의 추천시스템과는 통계적으로 유의한 성과 차이가 없었으나, 다른 두 모형에 대해서는 통계적으로 유의한 성과의 차이가 있는 것으로 나타났다.

• 전기학회논문지
• /
• 제67권4호
• /
• pp.569-577
• /
• 2018

In this paper, with the aid of information which is included within data, preprocessing algorithm-based black plastic classifier is designed. The slope and area of spectrum obtained by using laser induced breakdown spectroscopy(LIBS) are analyzed for each material and its ensuing information is applied as the input data of the proposed classifier. The slope is represented by the rate of change of wavelength and intensity. Also, the area is calculated by the wavelength of the spectrum peak where the material property of chemical elements such as carbon and hydrogen appears. Using informations such as slope and area, input data of the proposed classifier is constructed. In the preprocessing part of the classifier, Principal Component Analysis(PCA) and fuzzy transform are used for dimensional reduction from high dimensional input variables to low dimensional input variables. Characteristic analysis of the materials as well as the processing speed of the classifier is improved. In the condition part, FCM clustering is applied and linear function is used as connection weight in the conclusion part. By means of Particle Swarm Optimization(PSO), parameters such as the number of clusters, fuzzification coefficient and the number of input variables are optimized. To demonstrate the superiority of classification performance, classification rate is compared by using WEKA 3.8 data mining software which contains various classifiers such as Naivebayes, SVM and Multilayer perceptron.

• Molecular & Cellular Toxicology
• /
• 제3권1호
• /
• pp.60-67
• /
• 2007

Gentamicin is a broad-spectrum aminoglycoside antibiotic used in the treatment of bacterial infection. Although side effects of gentamicin such as nephrotoxicity and ototoxicity have been investigated, the information on the hepatic effects of gentamicin is still limited. In the present study, gene expression profiles were analyzed in the liver of gentamicin treated mice using Affymetrix GeneChip$^{(R)}$ Mouse Expression 430A 2.0 Array. Totally, 400 genes were identified as being either up- or down-regulated over 1.5-fold changes (P<0.01) in the liver of gentamicin treated mice. Among these deregulated genes, 16 up-regulated genes mainly involved in transport (Kif5b, Pex14, Rab14, Clcn3, and Necap1) and 20 down-regulated genes involved in lipid and other metabolisms (Hdlbp, Gm2a, Uroc1, and Dak) were selected using k-means clustering algorithm. The functional classification of differentially expressed genes represented that several stress-related genes were regulated in the liver by gentamicin treatment. This data may contribute in understanding the molecular mechanism in the liver of gentamicin treated mice.

• 대한산업공학회지
• /
• 제28권3호
• /
• pp.331-343
• /
• 2002

This paper proposes a fuzzy rule-based model as a means to build usability models between emotional satisfaction and design variables of consumer products. Based on a subtractive clustering algorithm, this model obtains partially overlapping rules from existing data and builds multiple local models each of which has a form of a linear regression equation. The best subset procedure and cross validation technique are used to select appropriate input variables. The proposed technique was applied to the modeling of luxuriousness, balance, and attractiveness of office chairs. For comparison, regression models were built on the same data in two different ways; one using only potentially important variables selected by the design experts, and the other using all the design variables available. The results showed that the fuzzy rule-based model had a great benefit in terms of the number of variables included in the model. They also turned out to be adequate for predicting the usability of a new product. Better yet, the information on the product classes and their satisfaction levels can be obtained by interpreting the rules. The models, when combined with the information from the regression models, are expected to help the designers gain valuable insights in designing a new product.

• 유통과학연구
• /
• 제20권6호
• /
• pp.1-10
• /
• 2022

Purpose: This research proposes a new modified Recency-Frequency-Monetary (RFM) model by extending the model with spatial analysis for supporting decision-makers in discovering the promotional target market. Research design, data and methodology: This quantitative research utilizes data-mining techniques and the RFM model to cluster a university's provider schools. The RFM model was modified by adapting its variables to the university's marketing context and adding a district's potential (D) variable based on heatmap analysis using Geographic Information System (GIS) and K-means clustering. The K-prototype algorithm and the Elbow method were applied to find provider school clusters using the proposed RFM-D model. After profiling the clusters, the target segment was assigned. The model was validated using empirical data from an Indonesian university, and its performance was compared to the Customer Lifetime Value (CLV)-based RFM utilizing accuracy, precision, recall, and F1-score metrics. Results: This research identified five clusters. The target segment was chosen from the highest-value and high-value clusters that comprised 17.80% of provider schools but can contribute 75.77% of students. Conclusions: The proposed model recommended more targeted schools in higher-potential districts and predicted the target segment with 0.99 accuracies, outperforming the CLV-based model. The empirical findings help university management determine the promotion location and allocate resources for promotional information distribution and student recruitment.

• 전자공학회논문지
• /
• 제52권6호
• /
• pp.117-125
• /
• 2015

앉아있는 시간이 긴 현대인들에게 바른 자세를 유지하도록 하는 것은 중요하다. 자세 교정을 위한 치료는 많은 시간과 비용이 소요되며, 전문의의 지속적인 관찰이 필요하다. 그러므로 사용자 스스로 자신의 자세를 판단하고 교정하기 위한 시스템이 필요하다. 본 논문에서는 사용자의 자세 데이터를 취득하여 취득된 자세가 정상자세인지 비정상자세인지 판단한다. 사용자의 자세 데이터 취득을 위해 관성 센서를 이용한 4개 관절 기반 모션캡쳐 시스템을 제안한다. 이 시스템을 통해 대상자의 자세 데이터를 취득하고, 취득한 데이터를 기반으로 특징을 추출하여 DB를 구축한다. 구축한 DB를 K-means 클러스터링 알고리즘을 이용하여 자세 학습을 수행한 후, 정상자세와 비정상자세를 분류한다. 관절의 회전각도, 위치정보, 분석정보를 이용하여 자세분류를 수행한 결과, 정상자세 판단 성공률은 99.79%로 나타났다. 이 결과로 미루어 4개 관절에 대한 특징을 이용하여 사용자의 자세를 판단 가능하며, 향후 척추질환 예방 시스템에 적용하여 사용자의 자세를 교정하는 데 도움을 줄 수 있을 것으로 판단된다.