• 제목/요약/키워드: adaptive model selection

검색결과 100건 처리시간 0.021초

자연 프루닝과 베이시안 선택에 의한 신경회로망 일반화 성능 향상 (Improving Generalization Performance of Neural Networks using Natural Pruning and Bayesian Selection)

  • 이현진;박혜영;이일병
    • 한국정보과학회논문지:소프트웨어및응용
    • /
    • 제30권3_4호
    • /
    • pp.326-338
    • /
    • 2003
  • 신경회로망 설계 및 모델선택의 목표는 최적의 구조를 가지는 일반화 성능이 우수한 네트워크를 구성하는 것이다. 하지만 학습데이타에는 노이즈(noise)가 존재하고, 그 수도 충분하지 않기 때문에 최종적으로 표현하고자 하는 진확률 분포와 학습 데이타에 의해 표현되는 경험확률분포(empirical probability density) 사이에는 차이가 발생한다. 이러한 차이 때문에 신경회로망을 학습데이타에 대하여 과다하게 적합(fitting)시키면, 학습데이타만의 확률분포를 잘 추정하도록 매개변수들이 조정되어 버리고, 진확률 분포로부터 멀어지게 된다. 이러한 현상을 과다학습이라고 하며, 과다학습된 신경회로망은 학습데이타에 대한 근사는 우수하지만, 새로운 데이타에 대한 예측은 떨어지게 된다. 또한 신경회로망의 복잡도가 증가 할수록 더 많은 매개변수들이 노이즈에 쉽게 적합되어 과다학습 현상은 더욱 심화된다. 본 논문에서는 통계적인 관점을 바탕으로 신경회로망의 일반화 성능을 향상시키는 신경회로 망의 설계 및 모델 선택의 통합적인 프로세스를 제안하고자 한다. 먼저 학습의 과정에서 적응적 정규화가 있는 자연기울기 학습을 통해 수렴속도의 향상과 동시에 과다학습을 방지하여 진확률 분포에 가까운 신경회로망을 얻는다. 이렇게 얻어진 신경회로망에 자연 프루닝(natural pruning) 방법을 적용하여 서로 다른 크기의 후보 신경회로망 모델을 얻는다. 이러한 학습과 복잡도 최적화의 통합 프로세스를 통하여 얻은 후보 모델들 중에서 최적의 모델을 베이시안 정보기준에 의해 선택함으로써 일반화 성능이 우수한 최적의 모델을 구성하는 방법을 제안한다. 또한 벤치마크 문제를 이용한 컴퓨터 시뮬레이션을 통하여, 제안하는 학습 및 모델 선택의 통합프로세스의 일반화 성능과 구조 최적화 성능의 우수성을 검증한다.

Joint Radio Selection and Relay Scheme through Optimization Model in Multi-Radio Sensor Networks

  • Lee, HyungJune
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제8권12호
    • /
    • pp.4451-4466
    • /
    • 2014
  • We present joint radio selection and relay scheme that delivers data from a source to a sink in heterogeneous stationary sensor networks consisting of various radio interfaces. The proposed scheme finds the optimal relay nodes and their corresponding radio interfaces that minimize energy consumption throughout the network while satisfying the end-to-end packet deadline requirement. We formulate the problem of routing through radio interface selection into binary integer programs, and obtain the optimal solution by solving with an optimization solver. We examine a trade-off relationship between energy consumption and packet delay based on network level simulations. We show that given the end-to-end deadline requirement, our routing algorithm finds the most energy-efficient routing path and radio interface across mesh hops. We demonstrate that the proposed routing scheme exploits the given packet delivery time to turn into network benefit of reducing energy consumption compared to routing based on single radio interface.

배경분리를 위한 개선된 적응적 가우시안 혼합모델에서의 동적 학습률 제어 (Dynamic Control of Learning Rate in the Improved Adaptive Gaussian Mixture Model for Background Subtraction)

  • 김영주
    • 한국정보통신학회:학술대회논문집
    • /
    • 한국해양정보통신학회 2005년도 추계종합학술대회
    • /
    • pp.366-369
    • /
    • 2005
  • 연속 영상을 이용하여 실시간으로 움직임 객체를 추출하고 추적하기 위해 배경분리(Background Subtraction) 기법을 주로 사용한다. 외부 환경에서는 조명의 변화, 나무의 흔들림과 같은 반복적인 움직임 그리고 급격히 움직이는 객체 등과 같이 고려해야할 많은 환경 변화 요인들이 존재한다. 이러한 외부 환경의 변화를 적응적으로 반영하여 배경을 분리할 수 있는 배경 모델로는 주로 가우시안 혼합 모델(GMM: Gaussian Mixture Model)이 적용되고 있으며, 실시간 성능 등을 개선시킨 적응적 가우시안 혼합 모델 등이 사용되고 있다. 본 논문은 개선된 적응적 가우시안 혼합 모델을 적용하고 고정된 학습률 ${\alpha}$(일반적으로 작은 값)을 사용함으로써 물체의 갑작스러운 움직임 등에 빠르게 적응하지 못하는 문제점을 해결하기 위해 가우시안 분포 수의 적응적 조절 기능과 픽셀 값을 분산을 이용하여 학습률 ${\alpha}$값을 동적으로 제어하는 방법을 제안하고 성능을 평가하였다.

  • PDF

Regularization Parameter Selection for Total Variation Model Based on Local Spectral Response

  • Zheng, Yuhui;Ma, Kai;Yu, Qiqiong;Zhang, Jianwei;Wang, Jin
    • Journal of Information Processing Systems
    • /
    • 제13권5호
    • /
    • pp.1168-1182
    • /
    • 2017
  • In the past decades, various image regularization methods have been introduced. Among them, total variation model has drawn much attention for the reason of its low computational complexity and well-understood mathematical behavior. However, regularization parameter estimation of total variation model is still an open problem. To deal with this problem, a novel adaptive regularization parameter selection scheme is proposed in this paper, by means of using the local spectral response, which has the capability of locally selecting the regularization parameters in a content-aware way and therefore adaptively adjusting the weights between the two terms of the total variation model. Experiment results on simulated and real noisy image show the good performance of our proposed method, in visual improvement and peak signal to noise ratio value.

Adaptive ridge procedure for L0-penalized weighted support vector machines

  • Kim, Kyoung Hee;Shin, Seung Jun
    • Journal of the Korean Data and Information Science Society
    • /
    • 제28권6호
    • /
    • pp.1271-1278
    • /
    • 2017
  • Although the $L_0$-penalty is the most natural choice to identify the sparsity structure of the model, it has not been widely used due to the computational bottleneck. Recently, the adaptive ridge procedure is developed to efficiently approximate a $L_q$-penalized problem to an iterative $L_2$-penalized one. In this article, we proposed to apply the adaptive ridge procedure to solve the $L_0$-penalized weighted support vector machine (WSVM) to facilitate the corresponding optimization. Our numerical investigation shows the advantageous performance of the $L_0$-penalized WSVM compared to the conventional WSVM with $L_2$ penalty for both simulated and real data sets.

모바일 장치와 공용 디스플레이 장치를 포함하는 환경에서 사용자의 특성에 기반한 Adaptive UI 설계 방안 (Method for Designing Adaptive UI Based on User's Context in the Environment Including Mobile Device and Public Display Device)

  • 강승수;고현;윤희용
    • 한국IT서비스학회지
    • /
    • 제11권4호
    • /
    • pp.181-194
    • /
    • 2012
  • The one of the most meaningful change in the recent ubiquitous environment is the omnipresence of public digital display device for providing ubiquitous information. It is the important issue to provide publicity as well as adaptive information to each user in the field of the public digital display device. This research proposes the idea ensuring fast response speed by the selection of user preference function. The preference function is selected by statistics using Zipf distribution in the system comprising mobile device and digital display device based on NFC (Near Field Communication). The idea is proved by CPM-GOMS model and the improvement of user response can be achieved.

Stable activation-based regression with localizing property

  • Shin, Jae-Kyung;Jhong, Jae-Hwan;Koo, Ja-Yong
    • Communications for Statistical Applications and Methods
    • /
    • 제28권3호
    • /
    • pp.281-294
    • /
    • 2021
  • In this paper, we propose an adaptive regression method based on the single-layer neural network structure. We adopt a symmetric activation function as units of the structure. The activation function has a flexibility of its form with a parametrization and has a localizing property that is useful to improve the quality of estimation. In order to provide a spatially adaptive estimator, we regularize coefficients of the activation functions via ℓ1-penalization, through which the activation functions to be regarded as unnecessary are removed. In implementation, an efficient coordinate descent algorithm is applied for the proposed estimator. To obtain the stable results of estimation, we present an initialization scheme suited for our structure. Model selection procedure based on the Akaike information criterion is described. The simulation results show that the proposed estimator performs favorably in relation to existing methods and recovers the local structure of the underlying function based on the sample.

개선된 IAFC 모델을 이용한 영상 대비 향상 기법 (An Image Contrast Enhancement Technique Using the Improved Integrated Adaptive Fuzzy Clustering Model)

  • 이금분;김용수
    • 한국지능시스템학회논문지
    • /
    • 제11권9호
    • /
    • pp.777-781
    • /
    • 2001
  • 본 논문은 저대비 영상을 처리하여 보다 향상된 영상을 얻고자 펴지 함소와 개선된 IAFC 모델을 적용한 영상 대비 향상 기법을 제안한다. 저대비에 의한 영상 정보의 불확실성이 무작위성보다 명암도의 모호성과 퍼지성에 근거한다는 점에서 퍼지 집합이론을 영상 향상 기법을 개발하는데 적용한다. 영상 향상의 단계를 퍼지화, 대비 강화 연산, 비퍼지화 단계로 나눠볼 수 있으며, 퍼지화 및 비퍼지화 과정에서 적절한 교차점 선택이 요구되고 이때 개선된 IAFC 모델을 적용하여 최적의 교차점을 선택한다. 데이터 대한 정신없이 임계 파라미터를 조정함으로써 클러스터링을 할 수 있는 개선된 IAFC 모델로 두 클래스만을 형성하도록 하여 명암도의 애매성이 최대가 되는 교차점을 찾아 대비를 강화시킨다. 대비 향상의 정략적 측정을 위해 퍼지성 지수를 사용하며 히스토그램 균등화 기법을 사용한 대비 향상 결과와 비교한다. 저대비 영상에 대해 최적의 교차점의 위치를 정하는 제안한 기법의 결과가 많은 실험영상을 통해 우수함을 보여주고 있다.

  • PDF

Software Effort Estimation in Rapidly Changing Computng Environment

  • Eung S. Jun;Lee, Jae K.
    • 한국지능정보시스템학회:학술대회논문집
    • /
    • 한국지능정보시스템학회 2001년도 The Pacific Aisan Confrence On Intelligent Systems 2001
    • /
    • pp.133-141
    • /
    • 2001
  • Since the computing environment changes very rapidly, the estimation of software effort is very difficult because it is not easy to collect a sufficient number of relevant cases from the historical data. If we pinpoint the cases, the number of cases becomes too small. However is we adopt too many cases, the relevance declines. So in this paper we attempt to balance the number of cases and relevance. Since many researches on software effort estimation showed that the neural network models perform at least as well as the other approaches, so we selected the neural network model as the basic estimator. We propose a search method that finds the right level of relevant cases for the neural network model. For the selected case set. eliminating the qualitative input factors with the same values can reduce the scale of the neural network model. Since there exists a multitude of combinations of case sets, we need to search for the optimal reduced neural network model and corresponding case, set. To find the quasi-optimal model from the hierarchy of reduced neural network models, we adopted the beam search technique and devised the Case-Set Selection Algorithm. This algorithm can be adopted in the case-adaptive software effort estimation systems.

  • PDF

Adaptive Importance Channel Selection for Perceptual Image Compression

  • He, Yifan;Li, Feng;Bai, Huihui;Zhao, Yao
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제14권9호
    • /
    • pp.3823-3840
    • /
    • 2020
  • Recently, auto-encoder has emerged as the most popular method in convolutional neural network (CNN) based image compression and has achieved impressive performance. In the traditional auto-encoder based image compression model, the encoder simply sends the features of last layer to the decoder, which cannot allocate bits over different spatial regions in an efficient way. Besides, these methods do not fully exploit the contextual information under different receptive fields for better reconstruction performance. In this paper, to solve these issues, a novel auto-encoder model is designed for image compression, which can effectively transmit the hierarchical features of the encoder to the decoder. Specifically, we first propose an adaptive bit-allocation strategy, which can adaptively select an importance channel. Then, we conduct the multiply operation on the generated importance mask and the features of the last layer in our proposed encoder to achieve efficient bit allocation. Moreover, we present an additional novel perceptual loss function for more accurate image details. Extensive experiments demonstrated that the proposed model can achieve significant superiority compared with JPEG and JPEG2000 both in both subjective and objective quality. Besides, our model shows better performance than the state-of-the-art convolutional neural network (CNN)-based image compression methods in terms of PSNR.