• Journal of Electrical Engineering and Technology
• /
• v.5 no.3
• /
• pp.363-370
• /
• 2010

Load forecasting has always been essential to the operation and planning of power systems in deregulated electricity markets. Various methods have been proposed for load forecasting, and the neural network is one of the most widely accepted and used techniques. However, to obtain more accurate results, more information is needed as input variables, resulting in huge computational costs in the learning process. In this paper, to reduce training time in multi-layer perceptron-based short-term load forecasting, a graphics processing unit (GPU)-based computing method is introduced. The proposed approach is tested using the Korea electricity market historical demand data set. Results show that GPU-based computing greatly reduces computational costs.

• Journal of Korean Society for Geospatial Information Science
• /
• v.4 no.2 s.8
• /
• pp.137-147
• /
• 1996

In this paper, the performances between maximum likelihood classifier based on statistical classification and multilayer perceptrons based on neural network approaches were compared and evaluated Experimental results from both neural network method and statistical method are presented. In addition, the nature of two different approches are analyzed based on the experiments.

• Journal of Sensor Science and Technology
• /
• v.11 no.4
• /
• pp.239-246
• /
• 2002

In this paper a new design of multi-layer perceptron(MLP) for camera calibration is proposed. Most existing techniques determine a transformation from 3D spatial points to their image points and camera parameters are tried to be estimated from the transformation. The technique proposed here, on the other hand, determines rays of sight uniquely from image points and parameters are estimated from the relationship using an MLP. By this approach projection and back-projection can be done more straightforwardly. Being based on a geometric model, a network design process becomes less ambiguous, which is a clear merit compared to other neural net based techniques. An MLP designed according to the technique proposed showed fast and stable learning in tests under various conditions.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2002.05a
• /
• pp.632-638
• /
• 2002

In novelty detection, one attempts to discriminate abnormal patterns from normal ones. Novelty detection is quite difficult since, unlike usual two class classification problems, only normal patterns are available for training. Auto-Associative Multi-Layer Perceptron (AAMLP) has been shown to provide a good performance based upon the property that novel patterns usually have larger auto-associative errors. In this paper, we give a mathematical analysis of 2-layer AAMLP's output characteristics and empirical results of 2-layer and 4-layer AAMLPs. Various activation functions such as linear, saturated linear and sigmoid are compared. The 2-layer AAMLPs cannot identify non-linear boundaries while the 4-layer ones can. When the data distribution is multi-modal, then an ensemble of AAMLPs, each of which is trained with pre-clustered data is required. This paper contributes to understanding of AAMLP networks and leads to practical recommendations regarding its use.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.6
• /
• pp.93-99
• /
• 1993

In previous paper we proposed a multi-layer bidirectional associative memory (MBAM) which is an extended model of the bidirectional associative memory (BAM) into a multilayer architecture. And we showed that the MBAM has the possibility to have binary storage for easy implementation. In this paper we present a MOdified MBAM(MOMBAM) with high performance compared to MBAM and multi-layer perceptron. The contents will include the architecture, the learning method, the computer simulation results for MOMBAM with MBAM and multi-layer perceptron, and the convergence properties shown by computer simulation examples.. And we will show that the proposed model can be used as classifier with a little restriction.

• Proceedings of the KIEE Conference
• /
• 2000.07d
• /
• pp.2977-2979
• /
• 2000

A Fuzzy Hybrid-Multilayer Perceptron (FH-MLP) Structure is proposed in this paper. proposed FH-MLP is not a fixed architecture. that is to say. the number of layers and the number of nodes in each layer of FH-MLP can be generated to adapt to the changing environment. FH-MLP consists of two parts. one is fuzzy nodes which each node is operated as a small fuzzy system with fuzzy implication rules. and its fuzzy system operates with Gaussian or Triangular membership functions in premise part and constants or regression polynomial equation in consequence part. the other is polynomial nodes which several types of high-order polynomial such as linear. quadratic. and cubic form are used and is connected as various kinds of multi-variable inputs. To demonstrate the effectiveness of the proposed method. time series data for gas furnace process has been applied.

• Journal of the Korea Society of Computer and Information
• /
• v.21 no.1
• /
• pp.1-8
• /
• 2016

Convolutional neural network (CNN) consists of a few pairs of both convolution layer and subsampling layer. Thus it has more hidden layers than multi-layer perceptron. With the increased layers, the size of convolution mask ultimately determines the total number of weights in CNN because the mask is shared among input images. It also is an important learning factor which makes or breaks CNN's learning. Therefore, this paper proposes the best method to choose the convolution size and the number of layers for learning CNN successfully. Through our face recognition with vast learning examples, we found that the best size of convolution mask is 5 by 5 and 7 by 7, regardless of the number of layers. In addition, the CNN with two pairs of both convolution and subsampling layer is found to make the best performance as if the multi-layer perceptron having two hidden layers does.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• v.9 no.1
• /
• pp.1133-1136
• /
• 2005

The MPEG/audio standard results from three years of co-work by an international committe of high-fidelity audio compression experts in the Moving Picture Experts Group (MPEG/audio). The MPEG standard is rigid only where necessary to ensure interoperability. In this paper, a new approach of training and classification of acoustic signal is addressed. This is some what a fields of application aspects rather than technonical problems such as MPEG/codec, MIDI. In preprocessing, acoustic signal is transformmed using DWT so as to extract a feature parameters of sound such as loudness, pitch, bandwidth and harmonicity. these accoustic parameters are exploited to the input vector of neural perceptron. Experimental results showed that proposed approach can be used for tunning the dissonance chord.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.10
• /
• pp.3098-3104
• /
• 2000

This paper defiens a perceptron-like self-organizing map(PSOM) and show that PSOM is equivalent to Kohonen's self-organizing map(SOM) if target values of output neurons of PSOM are selected properly. This fact imphes that PSOM is a generalized SOM algorithm. This paper also show that if clustering is restricted to vector sets distributed on hypersphere with unit radius, SOM and dot-product SOM(DOSM) are equivalent algorithms. Therefore we conclude that DSOM is a special case of SOM, which in turn a special, case of PSOM.

• Earthquakes and Structures
• /
• v.16 no.6
• /
• pp.679-690
• /
• 2019

This paper presents machine learning methods using Support Vector Machine (SVM) and Multilayer Perceptron (MLP) to analyze optimal damper distribution for structural vibration control. Regarding different building structures, a genetic algorithm based optimization method is used to determine optimal damper distributions that are further used as training samples. The structural features, the objective function, the number of dampers, etc. are used as input features, and the distribution of dampers is taken as an output result. In the case of a few number of damper distributions, multi-class prediction can be performed using SVM and MLP respectively. Moreover, MLP can be used for regression prediction in the case where the distribution scheme is uncountable. After suitable post-processing, good results can be obtained. Numerical results show that the proposed method can obtain the optimized damper distributions for different structures under different objective functions, which achieves better control effect than the traditional uniform distribution and greatly improves the optimization efficiency.