• Journal of KIBIM
• /
• v.9 no.1
• /
• pp.42-51
• /
• 2019

Maintenance of aging structures has attracted societal attention. Maintenance of the aging structure can be efficiently performed with a digital twin. In order to maintain the structure based on the digital twin, it is required to accurately detect the damage of the structure. Meanwhile, deep learning-based damage detection approaches have shown good performance for detecting damage of structures. However, in order to develop such deep learning-based damage detection approaches, it is necessary to use a large number of data before and after damage, but there is a problem that the amount of data before and after the damage is unbalanced in reality. In order to solve this problem, this study proposed a method based on Generative adversarial network, one of Generative Model, for generating acceleration data usually used for damage detection approaches. As results, it is confirmed that the acceleration data generated by the GAN has a very similar pattern to the acceleration generated by the simulation with structural analysis software. These results show that not only the pattern of the macroscopic data but also the frequency domain of the acceleration data can be reproduced. Therefore, these findings show that the GAN model can analyze complex acceleration data on its own, and it is thought that this data can help training of the deep learning-based damage detection approaches.

• Journal of KIISE:Software and Applications
• /
• v.29 no.8
• /
• pp.573-583
• /
• 2002

Pattern recognition process is made up of the feature extraction in the pre-processing, the pattern clustering by training and the recognition process. This paper presents the F-LVQ (Fuzzy Learning Vector Quantization) pattern classification model which is optimized by the fuzzy membership function for the OCR(Optical Character Recognition) system. We trained 220 numeric patterns of 22 Hangul and English fonts and tested 4840 patterns whose forms are changed variously. As a result of this experiment, it is proved that the proposed model is more effective and robust than other typical LVQ models.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.34S no.7
• /
• pp.102-111
• /
• 1997

In this paper, a patern recognition method of surface mount device(SMD) IC using wavelet transform and neural network is proposed. We chose the feature parameter according to the characteristics of coefficient matrix which is obtained from four level discrete wavelet transform (DWT). These feature parameters are normalized and then used for the input vector of neural network which is capable of adapting the surroundings such as variation of illumination, arrangement of objects and translation. Experimental results show that when the same form of feature pattern, as is used for learning, is put into neural network and gained 100% rate ofrecognition irrespective of SMD IC kinds, location and variation of illumination. In the case of unused feature pattern for learning, the recognition rate is 85.9% under the similar surroundings, where as an average recognition rate is 96.87% for the case of reregulated value of illumination. Proosed method is relatively simple compared with the traditional space domain method in extracting the feature parameter and is also well suited for recognizing the pattern's class, position and existence. It can also shorten the processing tiem better than method extracting feature parameter with the use of discrete cosine transform(DCT) and adapt the surroundings such as variation of illumination, the arrangement and the translation of SMD IC.

• Journal of Electrical Engineering and Technology
• /
• v.12 no.1
• /
• pp.363-372
• /
• 2017

This paper presents a novel face recognition algorithm based on the deep convolution neural network and key point detection jointed local binary pattern methodology to enhance the accuracy of face recognition. We firstly propose the modified face key feature point location detection method to enhance the traditional localization algorithm to better pre-process the original face images. We put forward the grey information and the color information with combination of a composite model of local information. Then, we optimize the multi-layer network structure deep learning algorithm using the Fisher criterion as reference to adjust the network structure more accurately. Furthermore, we modify the local binary pattern texture description operator and combine it with the neural network to overcome drawbacks that deep neural network could not learn to face image and the local characteristics. Simulation results demonstrate that the proposed algorithm obtains stronger robustness and feasibility compared with the other state-of-the-art algorithms. The proposed algorithm also provides the novel paradigm for the application of deep learning in the field of face recognition which sets the milestone for further research.

• The KIPS Transactions:PartB
• /
• v.11B no.4
• /
• pp.477-484
• /
• 2004

The technique of sequential pattern mining means generating a set of inter-transaction patterns residing in time-dependent data. This paper proposes a new method for generating sequential patterns with the use of Hellinger measure. While the current methods are generating single dimensional sequential patterns within a single attribute, the proposed method is able to detect multi-dimensional patterns among different attributes. A number of heuristics, based on the characteristics of Hellinger measure, are proposed to reduce the computational complexity of the sequential pattern systems. Some experimental results are presented.

• Fashion & Textile Research Journal
• /
• v.23 no.4
• /
• pp.480-490
• /
• 2021

Sales forecasting is crucial for many retail operations. For apparel retailers, accurate sales forecast for the next season is critical to properly manage inventory and plan their supply chains. The challenge in this increases because apparel products are always new for the next season, have numerous variations, short life cycles, long lead times, and seasonal trends. In this study, a sales forecasting model is proposed for apparel products using machine learning techniques. The sales data pertaining to outerwear items for four years were collected from a Korean sports brand and filtered with outliers. Subsequently, the data were standardized by removing the effects of exogenous variables. The sales patterns of outerwear items were clustered by applying K-means clustering, and outerwear attributes associated with the specific sales-pattern type were determined by using a decision tree classifier. Six types of sales pattern clusters were derived and classified using a hybrid model of clustering and decision tree algorithm, and finally, the relationship between outerwear attributes and sales patterns was revealed. Each sales pattern can be used to predict stock-keeping-unit-level sales based on item attributes.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.6B
• /
• pp.1120-1126
• /
• 2000

Recently, in order to analyze the time series problems that occur in the nature word, and analyzing method using a neural electric network is being studied more than a typical statistical analysis method. A neural electric network has a generalization performance that is possible to estimate and analyze about non-learning data through the learning of a population. In this paper, after collecting weather datum that was collected from 1987 to 1996 and learning a population established, it suggests the weather forecasting system for an estimation and analysis the future weather. The suggested weather forecasting system uses 28*30*1 neural network structure, raises the total learning numbers and accuracy letting the selecting concentration learning about the pattern, that is not collected, using the descending epsilon learning method. Also, the weather forecasting system, that is suggested through a comparative experiment of the typical time series analysis method shows more superior than the existing statistical analysis method in the part of future estimation capacity.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.85.4-85
• /
• 2001

A learning controller is presented for repetitive walking motion of biped robot. The learning control scheme learns the approximate inverse dynamics input of biped walking robot and uses the learned input pattern to generate an input profile of different walking motion from that learnt. In the learning controller, the PID feedback controller takes part in stabilizing the transient response of robot dynamics while the feedforward learning controller plays a role in computing the desired actuator torques for feedforward nonlinear dynamics compensation in steady state. It is shown that all the error signals in the learning control system are bounded and the robot motion trajectory converges to the desired one asymptotically. The proposed learning control scheme is ...

• Journal of the Korean Data and Information Science Society
• /
• v.15 no.2
• /
• pp.477-483
• /
• 2004

One of the major paradigms for supervised learning in neural network community is back-propagation learning. The standard implementations of back-propagation learning are optimal under the assumptions of identical and independent Gaussian noise. In this paper, for regression function estimation, we introduce $\varepsilon-insensitive$ back-propagation learning algorithm, which corresponds to minimizing the least absolute error. We compare this algorithm with support vector machine(SVM), which is another $\varepsilon-insensitive$ supervised learning algorithm and has been very successful in pattern recognition and function estimation problems. For comparison, we consider a more realistic model would allow the noise variance itself to depend on the input variables.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.39 no.3
• /
• pp.88-98
• /
• 2002

The multilayer perceptron (MLP) has several advantages against other pattern recognition methods, and is expected to be used as the learning and recognizing speakers of speaker verification system. But because of the low learning speed of the error backpropagation (EBP) algorithm that is used for the MLP learning, the MLP learning requires considerable time. Because the speaker verification system must provide verification services just after a speaker's enrollment, it is required to solve the problem. So, this paper tries to make short of time required to enroll speakers with the MLP based speaker verification system, using the method of improving the EBP learning speed and the method of reducing background speakers which adopts the cohort speakers method from the existing speaker verification.