• Journal of the Korean Institute of Intelligent Systems
• /
• v.8 no.5
• /
• pp.98-106
• /
• 1998

This paper proposes an efficient learning algorithm for improving the training performance of the neural network. The proposed method improves the training performance by applying the backpropagation algorithm of a global optimization method which is a hybrid of a stochastic approximation and a conjugate gradient method. The approximate initial point for f a ~gtl obal optimization is estimated first by applying the stochastic approximation, and then the conjugate gradient method, which is the fast gradient descent method, is applied for a high speed optimization. The proposed method has been applied to the parity checking and the pattern classification, and the simulation results show that the performance of the proposed method is superior to those of the conventional backpropagation and the backpropagation algorithm which is a hyhrid of the stochastic approximation and steepest descent method.

• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.3
• /
• pp.241-246
• /
• 2016

Nowadays many people have an interest in facial expression and the behavior of people. These are human-robot interaction (HRI) researchers utilize digital image processing, pattern recognition and machine learning for their studies. Facial feature point detector algorithms are very important for face recognition, gaze tracking, expression, and emotion recognition. In this paper, a cascade facial feature point detector is used for finding facial feature points such as the eyes, nose and mouth. However, the detector has difficulty extracting the feature points from several images, because images have different conditions such as size, color, brightness, etc. Therefore, in this paper, we propose an algorithm using a modified cascade facial feature point detector using a convolutional neural network. The structure of the convolution neural network is based on LeNet-5 of Yann LeCun. For input data of the convolutional neural network, outputs from a cascade facial feature point detector that have color and gray images were used. The images were resized to $32{\times}32$. In addition, the gray images were made into the YUV format. The gray and color images are the basis for the convolution neural network. Then, we classified about 1,200 testing images that show subjects. This research found that the proposed method is more accurate than a cascade facial feature point detector, because the algorithm provides modified results from the cascade facial feature point detector.

• Education of Primary School Mathematics
• /
• v.1 no.2
• /
• pp.85-95
• /
• 1997

Recently, by the popularization of computers and the development of many kinds of information transmission software, the living pattern in business offices and in home-life have changed rapidly. Because of the great progression of today's science technology, the influence of social education on the children is larger than that of the traditional school.. By a rapid change in the social atmosphere, there are some people who insist the traditional school education system is of no use any more. There have been many calls for reform of traditional schooling and in particular there has been major rethinking of school mathematics. The initial demand for change in elementary school mathematics is because of the poor achievement of students. There are even more compelling reasons for change. For example today's science technology society requires a different mathematical literacy for its citizens than that of the past. The importance of problem-solving based on interest and progress is more important than just paper-pencil computation in elementary schools. And also the increasing information wave of today's society demands that the school accept the long-distance education which could not be imagined in the past. Taking account of this variety, school education in the future should willingly introduce and apply the open education system to keep pace with today's society. To accept society demands actively, today's schools are going to accept and apply the idea of the open education. In this viewpoint, the purpose of the paper is to analyze the causes of under-achievement in mathematics teaming, the directions of school mathematics education, the system of textbooks and the problems of teaching-learning programs and paper-pencil test.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.8
• /
• pp.232-237
• /
• 2013

A newly developed fall detection algorithm using the HMM (Hidden Markov Model) extracted from the video is introduced. To distinguish between the fall from personal difference fall pattern or the normal activities of daily living (ADL), HMM machine learning algorithm is used. For getting fall feature vector of video, the motion vector from the optical flow is applied to the PCA (Principal Component Analysis). The combination of the angle, ratio of long-short axis, velocity from results of PCA make the new fall feature parameters. These parameters were applied to the HMM and the results were compared and analyzed. Among the newly proposed various kinds of fall parameters, the angle of movement showed the best results. The results show that this parameter can distinguish various types of fall from ADLs with 91.5% sensitivity and 88.01% specificity.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.12 no.5
• /
• pp.243-253
• /
• 2012

An opinion mining technique which was developed from document classification in area of data mining now becomes a common interest in domestic as well as international industries. The core of opinion mining is to decide precisely whether an opinion document is a positive or negative one. Although many related approaches have been previously proposed, a classification accuracy was not satisfiable enough to applying them in practical applications. A opinion documents written in Korean are not easy to determine a polarity automatically because they often include various and ungrammatical words in expressing subjective opinions. Proposed in this paper is a new approach of classification of opinion documents, which considers only a frequency of word patterns and excludes the grammatical factors as much as possible. In proposed method, we express a document into a bag of words and then apply a learning algorithm using a frequency of word patterns, and finally decide the polarity of the document using a score function. Additionally, we also present the experiment results for evaluating the accuracy of the proposed method.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.39 no.1
• /
• pp.25-30
• /
• 2016

It is critical to forecast the maximum daily and monthly demand for power with as little error as possible for our industry and national economy. In general, long-term forecasting of power demand has been studied from both the consumer's perspective and an econometrics model in the form of a generalized linear model with predictors. Time series techniques are used for short-term forecasting with no predictors as predictors must be predicted prior to forecasting response variables and containing estimation errors during this process is inevitable. In previous researches, seasonal exponential smoothing method, SARMA (Seasonal Auto Regressive Moving Average) with consideration to weekly pattern Neuron-Fuzzy model, SVR (Support Vector Regression) model with predictors explored through machine learning, and K-means clustering technique in the various approaches have been applied to short-term power supply forecasting. In this paper, SARMA and intervention model are fitted to forecast the maximum power load daily, weekly, and monthly by using the empirical data from 2011 through 2013. $ARMA(2,\;1,\;2)(1,\;1,\;1)_7$ and $ARMA(0,\;1,\;1)(1,\;1,\;0)_{12}$ are fitted respectively to the daily and monthly power demand, but the weekly power demand is not fitted by AREA because of unit root series. In our fitted intervention model, the factors of long holidays, summer and winter are significant in the form of indicator function. The SARMA with MAPE (Mean Absolute Percentage Error) of 2.45% and intervention model with MAPE of 2.44% are more efficient than the present seasonal exponential smoothing with MAPE of about 4%. Although the dynamic repression model with the predictors of humidity, temperature, and seasonal dummies was applied to foretaste the daily power demand, it lead to a high MAPE of 3.5% even though it has estimation error of predictors.

• The Korean Journal of Applied Statistics
• /
• v.32 no.5
• /
• pp.671-681
• /
• 2019

Studies on predicting arrhythmia using machine learning have been actively conducted with increasing number of arrhythmia patients. Existing studies have predicted arrhythmia based on multivariate data of feature variables extracted from RR interval data at a specific time point. In this study, we consider that the pattern of the heart state changes with time can be important information for the arrhythmia prediction. Therefore, we investigate the usefulness of predicting the arrhythmia with multivariate time series data obtained by extracting and accumulating the multivariate vectors of the feature variables at various time points. When considering 1-nearest neighbor classification method and its ensemble for comparison, it is confirmed that the multivariate time series data based method can have better classification performance than the multivariate data based method if we select an appropriate time series distance function.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.49 no.4
• /
• pp.42-53
• /
• 2012

In pattern recognition and machine learning society, classification has been a classical problem and the most widely researched area. Adaptive boosting also known as Adaboost has been successfully applied to binary classification problems. It is a kind of boosting algorithm capable of constructing a strong classifier through a weighted combination of weak classifiers. On the other hand, the PCA and LDA algorithms are the most popular linear feature extraction methods used mainly for dimensionality reduction. In this paper, the combination of Adaboost and feature extraction methods is proposed for efficient classification of two class data. Conventionally, in classification problems, the roles of feature extraction and classification have been distinct, i.e., a feature extraction method and a classifier are applied sequentially to classify input variable into several categories. In this paper, these two steps are combined into one resulting in a good classification performance. More specifically, each projection vector is treated as a weak classifier in Adaboost algorithm to constitute a strong classifier for binary classification problems. The proposed algorithm is applied to UCI dataset and FRGC dataset and showed better recognition rates than sequential application of feature extraction and classification methods.

• Journal of the Korean earth science society
• /
• v.30 no.1
• /
• pp.81-95
• /
• 2009

The purpose of this study wasto analyze the factors that enhance their learning achievement in a fieldtrip environment. For this academic goal, we analyzed a pattern of fieldtrip-related perception and attitudes of 19 science-talented students who participated in the 2007 KESO winter school. As for the perception type, the result of analysis showed that the science-talented students understood a fieldtrip as an experimental inquiry from an inquiry perspective, and that their understanding about a fieldtrip was based on anthropocentrism, positivism and instrumentalism from a science philosophy perspective. Regarding theattitudes type, the result revealed that the purpose of the winter school was mainly to learn knowledge in earth science, and that there was a significant tendency for the participating students to become a future scientist more eagerly than their parents expected. Students' fieldtrip-related academic self-concept was mostly positive while the participants experienced both positive and negative emotions.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2012.10a
• /
• pp.898-902
• /
• 2012

A descriptor which is suitable for motion analysis by using the motion features of moving objects from the real time image sequence is proposed. To segment moving objects from the background, the background learning is performed. We extract motion trajectories of individual objects by using the sequence of the $1^{st}$ order moment of moving objects. The center points of each object are managed by linked list. The descriptor includes the $1^{st}$ order coordinates of moving object belong to neighbor of the per-defined position in grid pattern, the start frame number which a moving object appeared in the scene and the end frame number which it disappeared. A video retrieval by the proposed descriptor combining global and local feature is more effective than conventional methods which adopt a single feature among global and local features.