• 정보과학회 논문지
• /
• 제45권1호
• /
• pp.15-21
• /
• 2018

낮은 계수 표현(Low-Rank Representation, LRR) 기반 방법은 얼굴 클러스터링, 객체 검출 등의 여러 실제 응용에 널리 사용되고 있다. 이 방법은 그래프 기반 준지도 학습에서 그래프 구축에 사용할 경우 높은 예측 정확도를 확보할 수 있어 많이 사용된다. 그러나 LRR 문제를 해결하기 위해서는 알고리즘의 매 반복마다 데이터 수 크기의 정방행렬에 대해 특이값 분해를 수행하여야 하므로 계산 비효율적이다. 이를 해결하기 위해 속도를 향상시킨 발전된 LRR 방법을 제안한다. 이는 최근 발표된 Fast LRR(FaLRR)을 기반으로 하며, FaLRR이 속도는 빠르지만 실제로 분류 문제에서 성능이 낮은 것을 해결하기 위해 기반 최적화 목표에 추가 제약 조건을 도입하고 이를 최적화하는 방법을 제안한다. 실험을 통하여 제안 방법은 LRR보다 더 좋은 해를 빠르게 찾아냄을 확인할 수 있다. 또한, 동일한 해를 도출하는 방법을 찾아내기는 어렵지만 최소화하는 목표가 추가될 경우 더 좋은 결과를 나타내는 Fast MLRR(FaMLRR)을 제안한다.

• International Journal of Internet, Broadcasting and Communication
• /
• 제10권4호
• /
• pp.50-64
• /
• 2018

This paper describes a visual object detection approach utilizing ensemble based machine learning. Object detection methods employing 1D features have the benefit of fast calculation speed. However, for real image with complex background, detection accuracy and performance are degraded. In this paper, we propose an ensemble learning algorithm that combines a 1D feature classifier and 2D DNF (Disjunctive Normal Form) classifier to improve the object detection performance in a single input image. Also, to improve the computing efficiency and accuracy, we propose a feature selecting method to reduce the computing time and ensemble algorithm by combining the 1D features and 2D DNF features. In the verification experiments, we selected the Haar-like feature as the 1D image descriptor, and demonstrated the performance of the algorithm on a few datasets such as face and vehicle.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 1992년도 춘계학술대회 논문집
• /
• pp.380-384
• /
• 1992

The position control of a pneumatic cylinder suing low cost on-off valves is studied. The valve control band(VCB) was proposed toget fast response and toprevent solenoid valves from unnecessary switching at the beginning of response. A learning algorithm was used to compensate the nonlinearity and complexity in mathematical modelling of pheumatic on-off controlled positioning systems. In this algorithm, the desired performance index and modified learning rate, were proposed to improvespeed and convergence of learning control. It is shown experimentally that the proposed algorithm is robust to changes of system parameters: the setting time less than 1.0 sec and the error bound of .+-. 0.1 mm can be obtained. The effects of supply pressure, size of switching valves and the effect of multiple valves are discussed, and computer simulation onthe dynamic performances of the pneumatic system is also presented.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 1996년도 추계학술대회 논문집
• /
• pp.994-998
• /
• 1996

We propose an approach to integrating fuzzy logic control with RBF(Radial Basis Function) networks and show how the integrated network can be applied to multivariable self-organizing and self-learning fuzzy controller. Using the hybrid learning algorithm. To investigate its usefulness and performance, this controller is applied to a time-delayed process system. Simulation results show good control performance and fast convergency in hybrid loaming method.

• Journal of Communications and Networks
• /
• 제8권4호
• /
• pp.432-441
• /
• 2006

As the community antenna television (CATV) networks becomes ubiquitous, instead of constructing an entirely new broadband network infrastructure, it has emerged as one of the rapid and economic technologies to interconnecting heterogeneous network to provide broadband access to subscribers. How to support ubiquitous real-time multimedia applications, especially in a heavy traffic environment, becomes a critical issue in modern CATV networks. In this paper, we propose a time guaranteed and efficient upstream minislots allocation algorithm for supporting quality-of-service (QoS) traffic over data over cable service interface specification (DOCSIS) CATV networks to fulfill the needs of realtime interactive services, such as video telephony, video on demand (VOD), distance learning, and so on. The proposed adaptive fast expansion algorithm and the loading statistics with dynamic swapping algorithm have been shown to perform better than that of the multimedia cable network system (MCNS) DOCSIS.

• Journal of Information Processing Systems
• /
• 제17권2호
• /
• pp.385-398
• /
• 2021

The complexity of deep learning models affects the real-time performance of gesture recognition, thereby limiting the application of gesture recognition algorithms in actual scenarios. Hence, a residual learning neural network based on a deep convolutional neural network is proposed. First, small convolution kernels are used to extract the local details of gesture images. Subsequently, a shallow residual structure is built to share weights, thereby avoiding gradient disappearance or gradient explosion as the network layer deepens; consequently, the difficulty of model optimisation is simplified. Additional convolutional neural networks are used to accelerate the refinement of deep abstract features based on the spatial importance of the gesture feature distribution. Finally, a fully connected cascade softmax classifier is used to complete the gesture recognition. Compared with the dense connection multiplexing feature information network, the proposed algorithm is optimised in feature multiplexing to avoid performance fluctuations caused by feature redundancy. Experimental results from the ISOGD gesture dataset and Gesture dataset prove that the proposed algorithm affords a fast convergence speed and high accuracy.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2001년도 하계학술대회 논문집 D
• /
• pp.2726-2728
• /
• 2001

This paper describes an optimal synthesis method of binary neural network(BNN) for an approximation problem of a circular region using a newly proposed learning algorithm[7] Our object is to minimize the number of connections and neurons in hidden layer by using a Newly Expanded and Truncated Learning Algorithm(NETLA) for the multilayer BNN. The synthesis method in the NETLA is based on the extension principle of Expanded and Truncated Learning(ETL) and is based on Expanded Sum of Product (ESP) as one of the boolean expression techniques. And it has an ability to optimize the given BNN in the binary space without any iterative training as the conventional Error Back Propagation(EBP) algorithm[6] If all the true and false patterns are only given, the connection weights and the threshold values can be immediately determined by an optimal synthesis method of the NETLA without any tedious learning. Futhermore, the number of the required neurons in hidden layer can be reduced and the fast learning of BNN can be realized. The superiority of this NETLA to other algorithms was proved by the approximation problem of one circular region.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2015년도 춘계학술대회
• /
• pp.294-297
• /
• 2015

최근 기계학습에 기반을 둔 빅데이터 분석이 큰 관심을 받으면서 다양한 학문 분야에 기계 학습 방안들이 접목되고 있다. 그 대표적인 분야 중 하나로 농축산 분야를 들 수 있고 실제 다양한 기계학습 방안들이 농축산분야에 적용되고 있다. 하지만 농축산에서 활용되는 기계학습의 경우 대부분 농업분야의 기후예측 및 축산분야의 유전자 분석 쪽으로 연구가 집중되어있고, 가축의 생체 데이터를 활용한 기계학습 방안은 많은 연구가 이루어지지 않았다. 본 연구에서는 가축의 실시간 생체 데이터를 이용하여 문제가 발생한 개체를 조기에 발견하는 방안을 제안하였다. 제안 방안에서는 기댓값 최대화 알고리즘을 이용하여 단일 가축 개체들의 실시간 생체 데이터를 2개의 클러스터로 나누고 이 두 클러스터 사이즈의 변화를 통해서 이상 개체를 조기에 판단한다. 특히 단일 개체의 문제와 전염성 질병 여부를 나누어 판단하므로 구제역과 같은 전염성 질병의 경우 빠른 대응을 가능케 하여 국가적 손실을 줄일 수 있게 한다. 더불어 제안 방안은 측정 생체 데이터에 대한 통계적 정보 없이도 적응적으로 클러스터를 형성할 수 있으므로 축사 외부의 환경 요소에 의해서 생체 데이터의 통계적 특성이 변화는 상황에서도 적응적으로 동작할 수 있다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권11호
• /
• pp.3099-3120
• /
• 2023

The present fast-moving era brings a serious stress issue that affects elders and youngsters. Everyone has undergone stress factors at least once in their lifetime. Stress is more among youngsters as they are new to the working environment. whereas the stress factors for elders affect the individual and overall performance in an organization. Electroencephalogram (EEG) based stress level classification is one of the widely used methodologies for stress detection. However, the signal processing methods evolved so far have limitations as most of the stress classification models compute the stress level in a predefined environment to detect individual stress factors. Specifically, machine learning based stress classification models requires additional algorithm for feature extraction which increases the computation cost. Also due to the limited feature learning characteristics of machine learning algorithms, the classification performance reduces and inaccurate sometimes. It is evident from numerous research works that deep learning models outperforms machine learning techniques. Thus, to classify all the emotions based on stress level in this research work a hybrid deep learning algorithm is presented. Compared to conventional deep learning models, hybrid models outperforms in feature handing. Better feature extraction and selection can be made through deep learning models. Adding machine learning classifiers in deep learning architecture will enhance the classification performances. Thus, a hybrid convolutional neural network model was presented which extracts the features using CNN and classifies them through machine learning support vector machine. Simulation analysis of benchmark datasets demonstrates the proposed model performances. Finally, existing methods are comparatively analyzed to demonstrate the better performance of the proposed model as a result of the proposed hybrid combination.

• Communications for Statistical Applications and Methods
• /
• 제27권2호
• /
• pp.255-268
• /
• 2020

A mixture of multivariate canonical fundamental skew t-distribution (CFUST) has been of interest in various fields. In particular, interest in the unsupervised learning society is noteworthy. However, fitting the model via EM algorithm suffers from significant processing time. The main cause is due to the calculation of many multivariate t-cdfs (cumulative distribution functions) in E-step. In this article, we provide an approximate, but fast calculation method for the in univariate fashion, which is the product of successively conditional univariate t-cdfs with Taylor's first order approximation. By replacing all multivariate t-cdfs in E-step with the proposed approximate versions, we obtain the admissible results of fitting the model, where it gives 85% reduction time for the 5 dimensional skewness case of the Australian Institution Sport data set. For this approach, discussions about rough properties, advantages and limits are also presented.