• Proceedings of the KIEE Conference
• /
• 2005.11b
• /
• pp.207-210
• /
• 2005

The Similarity Index[1] is a good Performance measure for the network reduction. It can be applied to the network reduction In the zone categorized by the nodal prices. This paper deals with a zonal reduction method based on the similarity indices. The proposed method was verified by IEEE 39 bus test system.

• Communications for Statistical Applications and Methods
• /
• v.26 no.5
• /
• pp.497-506
• /
• 2019

Forecasting the U.S. employment level is made using machine learning methods of the artificial neural network: deep neural network, long short term memory (LSTM), gated recurrent unit (GRU). We consider the big data of the federal reserve economic data among which 105 important macroeconomic variables chosen by McCracken and Ng (Journal of Business and Economic Statistics, 34, 574-589, 2016) are considered as predictors. We investigate the influence of the two statistical issues of the dimension reduction and time series differencing on the machine learning forecast. An out-of-sample forecast comparison shows that (LSTM, GRU) with differencing performs better than the autoregressive model and the dimension reduction improves long-term forecasts and some short-term forecasts.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.12 no.20
• /
• pp.47-53
• /
• 1989

This paper presents an algorithm CAPFACT to evaluate the reliability of a capacitated two terminal network such as a communication network, a power distribution network, and a pipeline network. The network is good(working) if and only if it is possible to transmit successfully the required system capacity from one specified terminal to the other. This paper defines new Capacitated series-parallel reduction to be applied to a series-parallel structure of the network. New Capacitated factoring method is applied to a non-series-parallel structure. The method is based on the factoring theorem given by Agrawal and Barlow. According to the existing studies on the reliability evaluation of the network that the capacity is not considered, the factoring method using reduction is efficient. The CAPFACT is more efficient than Aggarwal algorithm which enumerated and combined the paths. The efficiency is proved by the result of testing the number of operations and cpu time on FORTRAN compiler of VAX-11/780 at Hanyang University.

• Phonetics and Speech Sciences
• /
• v.7 no.4
• /
• pp.3-8
• /
• 2015

In this paper, we investigate an input dimension reduction method using continuous word vector in deep neural network language model. In the proposed method, continuous word vectors were generated by using Google's Word2Vec from a large training corpus to satisfy distributional hypothesis. 1-of-${\left|V\right|}$ coding discrete word vectors were replaced with their corresponding continuous word vectors. In our implementation, the input dimension was successfully reduced from 20,000 to 600 when a tri-gram language model is used with a vocabulary of 20,000 words. The total amount of time in training was reduced from 30 days to 14 days for Wall Street Journal training corpus (corpus length: 37M words).

• Journal of information and communication convergence engineering
• /
• v.10 no.2
• /
• pp.162-167
• /
• 2012

This paper proposes a speech processing system based on a model of the human auditory system and a noise reduction neural network with fast Fourier transform (FFT) amplitude and phase spectrums for noise reduction under background noise environments. The proposed system reduces noise signals by using the proposed neural network based on FFT amplitude spectrums and phase spectrums, then implements auditory processing frame by frame after detecting voiced and transitional sections for each frame. The results of the proposed system are compared with the results of a conventional spectral subtraction method and minimum mean-square error log-spectral amplitude estimator at different noise levels. The effectiveness of the proposed system is experimentally confirmed based on measuring the signal-to-noise ratio (SNR). In this experiment, the maximal improvement in the output SNR values with the proposed method is approximately 11.5 dB better for car noise, and 11.0 dB better for street noise, when compared with a conventional spectral subtraction method.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.2
• /
• pp.832-854
• /
• 2019

Action recognition has been studied in computer vision field for years. We present an effective approach to recognize actions using a dimension reduction method, which is applied as a crucial step to reduce the dimensionality of feature descriptors after extracting features. We propose to use sparse matrix and randomized kd-tree to modify it and then propose modified Local Fisher Discriminant Analysis (mLFDA) method which greatly reduces the required memory and accelerate the standard Local Fisher Discriminant Analysis. For feature encoding, we propose a useful encoding method called mix encoding which combines Fisher vector encoding and locality-constrained linear coding to get the final video representations. In order to add more meaningful features to the process of action recognition, the convolutional neural network is utilized and combined with mix encoding to produce the deep network feature. Experimental results show that our algorithm is a competitive method on KTH dataset, HMDB51 dataset and UCF101 dataset when combining all these methods.

• Journal of the Korean Society for information Management
• /
• v.39 no.2
• /
• pp.233-254
• /
• 2022

Link reduction algorithms such as pathfinder network are the widely used methods to overcome problems with the visualization of weighted networks for knowledge domain analysis. This study proposed NetRSQ, an indicator to measure the goodness of fit of a link reduction algorithm for the network visualization. NetRSQ is developed to calculate the fitness of a network based on the rank correlation between the path length and the degree of association between entities. The validity of NetRSQ was investigated with data from previous research which qualitatively evaluated several network generation algorithms. As the primary test result, the higher degree of NetRSQ appeared in the network with better intellectual structures in the quality evaluation of networks built by various methods. The performance of 4 link reduction algorithms was tested in 40 datasets from various domains and compared with NetRSQ. The test shows that there is no specific link reduction algorithm that performs better over others in all cases. Therefore, the NetRSQ can be a useful tool as a basis of reliability to select the most fitting algorithm for the network visualization of intellectual structures.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.27 no.1
• /
• pp.33-51
• /
• 2002

A fully connected Artificial Neural Network (ANN) contains many connections. Compared to the pruned ANN with fewer connections, the fully connected ANN takes longer time to produce solutions end may not provide appropriate solutions to new unseen date. Therefore, by reducing the sloe of ANN, we can overcome the overfitting problem and increase the computing speed. In this research, we reduced the size of ANN by destroying the connections. In other words, we investigated the performance change of the reduced ANN by systematically destroying the connections. Then we found the acceptable level of connection-destruction on which the resulting ANN Performs as well as the original fully connected ANN. In the previous researches on the sloe reduction of ANN, the reduced ANN had to be retrained every time some connections were eliminated. Therefore, It tool lolly time to obtain the reduced ANN. In this research, however, we provide the acceptable level of connection-destruction according to the size of the fully connected ANN. Therefore, by applying the acceptable level of connection-destruction to the fully connected ANN without any retraining, the reduced ANN can be obtained efficiently.

• International Journal of Control, Automation, and Systems
• /
• v.4 no.3
• /
• pp.359-364
• /
• 2006

This article describes the useful way to measure the torque and RPM of the geared motor. For this we have made the planetary geared reduction motor including 2 Hall sensors in it and the monitoring system. The monitoring system displays the sensing values (torque, rpm) and the calculated value (power) and it also has the network capability using the Bluetooth protocol. We will show that our solution is much more inexpensive and simple method to measure torque and rpm than before.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.9 no.4
• /
• pp.216-224
• /
• 2009

The endless scaling down of the semiconductor process made the impact of the power network noise on the performance of the state-of-the-art chip a serious design problem. This paper compares the performances of two popular circuit reduction approaches used to improve the efficiency of power network noise analysis: moment matching-based model order reduction (MOR) and node elimination-based MOR. As the benchmarks, we chose PRIMA and R2Power as the matching-based MOR and the node elimination-based MOR. Experimental results indicate that the accuracy, efficiency, and memory requirement of both methods very strongly depend on the structure of the given circuit, i.e., numbers of the nodes and sources, and the number of moments to preserve for PRIMA. PRIMA has higher accuracy in general, while the error of R2Power is also in the acceptable range. On the other hand, PRIMA has the higher efficiency than R2Power, only when the numbers of nodes and sources are small enough. Otherwise, R2Power clearly outperforms PRIMA in efficiency. In the memory requirement, the memory size of PRIMA increases very quickly as the numbers of nodes, sources, and preserved moments increase.