• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제16권1호
• /
• pp.27-35
• /
• 2016

Black-box classifiers, such as artificial neural network and support vector machine, are a popular classifier because of its remarkable performance. They are applied in various fields such as inductive inferences, classifications, or regressions. However, by its characteristics, they cannot provide appropriate explanations how the classification results are derived. Therefore, there are plenty of actively discussed researches about interpreting trained black-box classifiers. In this paper, we propose a method to make a fuzzy logic-based classifier using extracted rules from the artificial neural network and support vector machine in order to interpret internal structures. As an object of classification, an anomalous propagation echo is selected which occurs frequently in radar data and becomes the problem in a precipitation estimation process. After applying a clustering method, learning dataset is generated from clusters. Using the learning dataset, artificial neural network and support vector machine are implemented. After that, decision trees for each classifier are generated. And they are used to implement simplified fuzzy logic-based classifiers by rule extraction and input selection. Finally, we can verify and compare performances. With actual occurrence cased of the anomalous propagation echo, we can determine the inner structures of the black-box classifiers.

• International Journal of Computer Science & Network Security
• /
• 제21권6호
• /
• pp.207-212
• /
• 2021

Clustering Unlabeled Spatial-datasets to convert them to Labeled Spatial-datasets is a challenging task specially for geographical information systems. In this research study we investigated the NYC Taxi Limousine Commission dataset and discover that all of the spatial-temporal trajectory are unlabeled Spatial-datasets, which is in this case it is not suitable for any data mining tasks, such as classification and regression. Therefore, it is necessary to convert unlabeled Spatial-datasets into labeled Spatial-datasets. In this research study we are going to use the Clustering Technique to do this task for all the Trajectory datasets. A key difficulty for applying machine learning classification algorithms for many applications is that they require a lot of labeled datasets. Labeling a Big-data in many cases is a costly process. In this paper, we show the effectiveness of utilizing a Clustering Technique for labeling spatial data that leads to a high-accuracy classifier.

• ETRI Journal
• /
• 제30권2호
• /
• pp.338-340
• /
• 2008

In this letter, we propose a novel approach to detecting and tracking apartment buildings for the development of a video-based navigation system that provides augmented reality representation of guidance information on live video sequences. For this, we propose a building detector and tracker. The detector is based on the AdaBoost classifier followed by hierarchical clustering. The classifier uses modified Haar-like features as the primitives. The tracker is a motion-adjusted tracker based on pyramid implementation of the Lukas-Kanade tracker, which periodically confirms and consistently adjusts the tracking region. Experiments show that the proposed approach yields robust and reliable results and is far superior to conventional approaches.

• 인지과학
• /
• 제23권3호
• /
• pp.295-322
• /
• 2012

본 연구에서는 벡터 공간 모델과 HAL (Hyperspace Analog to Language)을 적용해서 단어 의미 유사성을 군집한다. 일정한 크기의 문맥을 통해서 단어 간의 상관성을 측정하는 HAL을 도입하고(Lund and Burgess 1996), 상관성 측정에서 고빈도와 저빈도에 다르게 측정되는 왜곡을 줄이기 위해서 벡터 공간 모델을 적용해서 단어 쌍의 코사인 유사도를 측정하였다(Salton et al. 1975, Widdows 2004). HAL과 벡터 공간 모델로 만들어지는 공간은 다차원이므로, 차원을 축소하기 위해서 PCA (Principal Component Analysis)와 SVD (Singular Value Decomposition)를 적용하였다. 유사성 군집을 위해서 비감독 방식과 감독 방식을 적용하였는데, 비감독 방식에는 클러스터링을 감독 방식에는 SVM (Support Vector Machine), 나이브 베이즈 구분자(Naive Bayes Classifier), 최대 엔트로피(Maximum Entropy) 방식을 적용하였다. 이 연구는 언어학적 측면에서 Harris (1954), Firth (1957)의 분포 가설(Distributional Hypothesis)을 활용한 의미 유사도를 측정하였으며, 심리언어학적 측면에서 의미 기억을 설명하기 위한 모델로 벡터 공간 모델과 HAL을 결합하였으며, 전산적 언어 처리 관점에서 기계학습 방식 중 감독 기반과 비감독 기반을 적용하였다.

• Journal of Biosystems Engineering
• /
• 제33권5호
• /
• pp.340-347
• /
• 2008

Dipstick urinalysis is used as a routine test for a screening test of UTI (urinary tract infection) in primary practice because urine dipstick test is simple. The result of dipstick urinalysis brings medical professionals to make a microscopic examination and urine culture for exact UTI diagnosis, therefore it is emphasized on a role of screening test. The objective of this study was to the classification between UTI patients and normal subjects using hybrid neural network classifier with enhanced clustering performance in urine dipstick screening test. In order to propose a classifier, we made a hybrid neural network which combines with RBF layer, summation & normalization layer and L VQ artificial neural network layer. For the demonstration of proposed hybrid neural network, we compared proposed classifier with various artificial neural networks such as back-propagation, RBFNN and PNN method. As a result, classification performance of proposed classifier was able to classify 95.81% of the normal subjects and 83.87% of the UTI patients, total average 90.72% according to validation dataset. The proposed classifier confirms better performance than other classifiers. Therefore the application of such a proposed classifier expect to utilize telemedicine to classify between UTI patients and normal subjects in the future.

• 한국지능시스템학회논문지
• /
• 제25권3호
• /
• pp.242-247
• /
• 2015

본 논문은 방사형 기저함수 신경회로망(Radial Basis Function Neural Network) 패턴분류기를 기반으로 강수 에코와 비(非)강수 에코를 분류하는 방법을 제시한다. 강수 에코와 비(非)강수 에코를 분류하기 위하여 기상레이더 자료의 특성을 분석하였다. 이를 기반으로 UF 데이터의 전처리를 실시하여 입력변수(DZ, SDZ, VGZ, SPN, DZ_FR, VR)를 선정 하였고 학습데이터 및 테스트데이터로 구성하였다. 마지막으로, 기상청에서 사용되고 있는 QC 데이터는 제안된 알고리즘의 성능을 비교하기 위해 사용하였다.

• 전기학회논문지
• /
• 제64권7호
• /
• pp.1064-1073
• /
• 2015

In this study, we introduce the pedestrian detection system by using the feature of HOG-PCA and RBFNNs pattern classifier. HOG(Histogram of Oriented Gradient) feature is extracted from input image to identify and recognize a object. And a dimension is reduced for improving performance as well as processing speed by using PCA which is a typical dimensional reduction algorithm. So, the feature of HOG-PCA through the dimensional reduction by using PCA leads to the improvement of the detection rate. FCM clustering algorithm is used instead of gaussian function to apply the characteristic of input data as well and connection weight is used by polynomial expression such as constant, linear, quadratic and modified quadratic. Finally, INRIA person database known as one of the benchmark dataset used for pedestrian detection is applied for the performance evaluation of the proposed classifier. The experimental result of the proposed classifier are compared with those studied by Dalal.

• 한국의류학회지
• /
• 제44권6호
• /
• pp.1053-1069
• /
• 2020

This study examined the sales pattern relationship with respect to product attributes to propose sales forecasting for fashion products. We analyzed 537 SKU sales data of T-shirts in the domestic sports brand using SAS program. The sales pattern of fashion products fluctuated and were influenced by exogenous factors; therefore, we removed the influence of exogenous factors found to be price discounts and holiday effects as a result of regression analysis. In addition, it was difficult to predict sales using the sales patterns of the same product since fashion products were released as new products every year. Therefore, the forecasting model was proposed using sales patterns of related product attributes when attributes were considered descriptive variables. We classified sales patterns using K-means clustering in order to explain the relationship between sales patterns and product attributes along with creating a decision tree classifier using attributes as input and sales patterns as output. As a result, the sales patterns of T-shirts were clustered into six types that featured the characteristic shape of peak and slope. It was also associated with the combination of product attributes and their values in regards to the proposed sales pattern prediction model.

• Journal of information and communication convergence engineering
• /
• 제4권4호
• /
• pp.174-179
• /
• 2006

In this paper we present a text detection method inspired by wavelet packet analysis and improved fuzzy clustering algorithm(IAFC).This approach assumes that the text and non-text regions are considered as two different texture regions. The text detection is achieved by using wavelet packet analysis as a feature analysis. The wavelet packet analysis is a method of wavelet decomposition that offers a richer range of possibilities for document image. From these multi scale features, we adapt the improved fuzzy clustering algorithm based on the unsupervised learning rule. The results show that our text detection method is effective for document images scanned from newspapers and journals.

• 한국지능시스템학회논문지
• /
• 제13권4호
• /
• pp.409-415
• /
• 2003

Recently, the modular learning are very popular and receive much attention for pattern classification. The modular learning method based on the "divide and conquer" strategy can not only solve the complex problems, but also reach a better result than a single classifier′s on the learning quality and speed. In the neural network area, some researches that take the modular learning approach also have been made to improve classification performance. In this paper, we propose a simple modular neural network for characters recognition of vehicle number plate and evaluate its performance on the clustering methods of feature vectors used in constructing subnetworks. We implement two clustering method, one is grouping similar feature vectors by K-means clustering algorithm, the other grouping unsimilar feature vectors by our proposed algorithm. The experiment result shows that our algorithm achieves much better performance.