• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.2
• /
• pp.103-111
• /
• 2023

This paper proposes a new data correction technique that transforms anomalies in time series data into normal values. With the recent development of IT technology, a vast amount of time-series data is being collected through sensors. However, due to sensor failures and abnormal environments, most of time-series data contain a lot of anomalies. If we build a predictive model using original data containing anomalies as it is, we cannot expect highly reliable predictive performance. Therefore, we utilizes the LSTM-GAN model to detect anomalies in the original time series data, and combines DTW (Dynamic Time Warping) and GAN techniques to replace the anomaly data with normal data in partitioned window units. The basic idea is to construct a GAN model serially by applying the statistical information of the window with normal distribution data adjacent to the window containing the detected anomalies to the DTW so as to generate normal time-series data. Through experiments using open NAB data, we empirically prove that our proposed method outperforms the conventional two correction methods.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2012.05a
• /
• pp.369-370
• /
• 2012

일반적으로 회귀분석의 최적화는 평균적인 개념을 확장하여 사용되어지고 있다. 평균은 관찰값들에 관한 모든 정보와 관련된 통계량으로써 많은 연구에 이용되어지고 있다. 정규분포를 이루는 모집단의 경우 평균을 사용한 추정이 바람직하지만, 이상치로 인한 분포의 꼬리가 두꺼워지는 경우 중위수(median)를 사용하는 것이 바람직하다고 알려져 있다. 강수량의 분포형태는 꼬리(tail)가 두꺼운 왜곡된 형태를 갖고 있으므로 robust 통계량인 Quantile을 이용한 강수량의 분석 및 평가를 실시하였다. 본 연구에서는 Quantile에 따른 회귀선의 변화를 이용하여 강수량의 경향성을 평가하고, 극치강수량의 변화를 보여줄 수 있는 Quantle값을 추출해 보고자 한다. 또한 bootstrap 방법을 이용하여 Quantile에 따른 회귀계수의 신뢰구간을 분석하여 회귀인자의 신뢰성을 평가하였다. 본 연구에서 적용한 Quantile Regression 기법은 회귀계수의 추정에 있어서 회귀인자의 신뢰성을 Quantile-회귀계수 그래프를 통해 분석할 수 있으며, 이상값의 영향을 저감시키는 평균과 달리 이상값의 영향을 효과적으로 분리 및 재현시킬 수 있어 극치값에 따른 변화를 효과적으로 평가할 수 있으며, robust 통계량의 특징인 분산이 적은 안정적인 추정량을 확보할 수 있다.

• Economic and Environmental Geology
• /
• v.43 no.3
• /
• pp.249-258
• /
• 2010

In general, a crustal geomagnetic (or gravity) anomaly compiled at one altitude can be estimated at a different altitude by continuation using the Fourier transform (FT). However, in case of continuation with a great distance between the two elevations, or, in particular, in case of downward continuation, the estimated anomalies by the FT are likely to be mathematically unstable so that the estimated values are not realistic. To solve this problem, two independently measured magnetic field anomalies at different altitudes, such as aeromagnetic and satellite magnetic observations, are implemented to estimate values at in-between altitude for better understanding and interpreting geophysical and geological features. This ‘'dual continuation’' technique is straightforward in the FT and gives a more realistic estimate in all altitudes when we simulated with a set of prismatic bodies at different altitudes. This implies that we add up another constraint like satellite-based observations on the geopotential field modeling for the non-unique geological and geophysical problems to a conventional Fourier-type continuation technique with a single set of observations.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.1
• /
• pp.131-142
• /
• 2016

Outliers in surveys can have a large effect on estimates of totals. This is especially true in business surveys where the populations are drawn are typically skewed. In this paper, we discussed the practical development and implementation of methods to identify and deal with outliers. A detection method is based on quartile method and detected outlier is processed in various ways. The study examines two versions of winsorised estimators with three different cut-off thresholds for each one. For the simulation study, four types of weight transformation function have been considered.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2020.01a
• /
• pp.7-9
• /
• 2020

기존 CCTV 비디오에서 딥러닝 기반의 이상 탐지 연구는 객체의 행동 값만을 이용하여 이상을 탐지하였기 때문에, 시간 흐름에 따른 정보가 축소되는 문제점이 있었다. 그러나 CCTV 비디오에서의 이상의 원인은 다양한 요소와 시계열 분석에 따른 정보로 이루어져 있어 시간 정보를 유지하면서 다양한 특징 값을 사용한 모델을 설계할 필요가 있다. 따라서 본 논문에서는 C3D에 광학 흐름을 결합한 새로운 앙상블 모델을 제안한다. 실험 결과 본 논문에서 제안하는 모델이 75.83의 AUC를 얻어 기존에 연구되었던 행동 값만을 사용한 모델보다 높은 정확도를 달성하였다. 또한 이상 탐지 모델 설계 시 객체의 행동에 다양한 측면을 고려할 수 있는 여러 특징 값과 시계열 분석에 따른 정보를 사용하는 것이 적절하다는 결론을 도출하였다.

• Journal of IKEEE
• /
• v.25 no.4
• /
• pp.710-715
• /
• 2021

Power line signal can show disturbances due to various causes. Typical anomalies are temporary sag/swell of the amplitude, flat topped signal, and harmonic distortions. The disturbances need to be detected and treated properly for the quality of the power signal. In this study, the power disturbances are detected using the inflection points (IP). The inflection points are defined as points where local maxima/minima or the slope changes occur. The power line signal has a fixed IP pattern since it is basically sinusoidal, and it may have additional inflection points if there is any disturbance. The disturbance is detected by comparing the IP patterns between the normal signal and distorted signal. In addition, by defining a cost function, the time instant where the disturbance happens can be decided. The computer simulation shows that the proposed method is useful for the detection of various disturbances. The simple sag or swell signal only shows the amplitude changes at the detected inflection points. However, the flat top signal and harmonically distorted signal produce additional inflection points and large values in the cost function. These results can be exploited for the further processing of disturbance classification.

• Journal of Korean Ophthalmic Optics Society
• /
• v.14 no.3
• /
• pp.51-57
• /
• 2009

Purpose: The purpose of this study was to determine if monocular and binocular accommodative facility tests would be useful in diagnosing general binocular dysfunctions. Methods: 95 symptomatic children, who were selected from comprehensive vision tests, were classified into four groups (29 subjects with accommodative dysfunctions, 28 subjects with vergence dysfunctions, 25 subjects with combined accommodative and vergence dysfunctions, 13 subjects with normal binocular functions). Monocular and binocular accommodative facility was measured with ${\pm}$2.00 D flipper lenses. Results: Statistical analysis revealed that binocular accommodative facility measurement was significantly lower than monocular accommodative measurement in the vergence dysfunction group (p<0.01). However, there were no differences between monocular and binocular accommodative facility measurements in the group of accommodative or combined accommodative and vergence dysfunction (p>0.05). In addition, subjects with general binocular dysfunctions performed significantly poorer than subjects with normal binocular function on monocular and binocular accommodative facility (p<0.000). Conclusions: As a result of this study, monocular and binocular accommodative facility test, which could differentiate dysfunction from normal as well as between dysfunctions, indicated useful means for diagnosis of general binocular dysfunctions.

• The Korean Journal of Applied Statistics
• /
• v.20 no.3
• /
• pp.561-571
• /
• 2007

Local linear regression estimator is the most widely used nonparametric regression estimator which has a number of advantages over the traditional kernel estimators. It is well known that local linear estimator can produce erratic result in sparse regions in the realization of the design and the interpolation method of Hall and Turlach (1997) is the very efficient way to resolve this problem. However, it has been never pointed out that Hall and Turlach's interpolation method is very sensitive to outliers. In this paper, we propose the robust version of the interpolation method for adapting to sparse design. The finite sample properties of the method is compared with Hall and Turlach's method by the simulation study.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.29 no.5
• /
• pp.543-550
• /
• 2011

The gravity anomaly is a basic geophysical data applied in various fields such as geophysics, geodesy and national defense. In general, the gravity anomaly is used through a interpolation process based on the constructed database. The gravity variation, however, is appeared in various shapes depending on the topography and the density of the underground structures. Therefore, the interpolation could lead to a large differences if the gravity fields do not satisfy the assumptions on the signal behavior like linear or a certain degree polynomials. Furthermore, the interpolation does not reflect the physical characteristics of the gravity such as the harmonic condition. In this study, the gravity modeling using the plane Fourier series and radial basis functions are performed to overcome the problems in the usual interpolation. The results of the modeling is analyzed for the case of the gravity referenced navigation focused on the signal characteristics. Based on the study, it was found that the results from modeling are not much different to that from the interpolation in a smoothly varied area. In case of the highly varied area, however, a large differences are appeared among the three methods. Especially, the Fourier series shows the most smooth variations in the modeled gravity values while the highest variations appeared in the interpolation. Applying to the gravity referenced navigation, it was found that the modeling is more effective in calculation cost. It is considered that the results from this study provides a basis on effective modeling of the gravity fields in terms of the signal characteristics and resolution for various application fields.

• Korean Journal of Remote Sensing
• /
• v.28 no.4
• /
• pp.357-367
• /
• 2012

This paper presents a two-stage methodology for anomaly detection from hyperspectral imagery that consists of transform-based feature extraction and selection, and computation of a local spatial auto-correlation statistic. First, principal component transform and 3D wavelet transform are applied to reduce redundant spectral information from hyperspectral imagery. Then feature selection based on global skewness and the portion of highly skewed sub-areas is followed to find optimal features for anomaly detection. Finally, a local indicator of spatial association (LISA) statistic is computed to account for both spectral and spatial information unlike traditional anomaly detection methodology based only on spectral information. An experiment using airborne CASI imagery is carried out to illustrate the applicability of the proposed anomaly detection methodology. From the experiments, anomaly detection based on the LISA statistic linked with the selection of optimal features outperformed both the traditional RX detector which uses only spectral information, and the case using major principal components with large eigen-values. The combination of low- and high-frequency components by 3D wavelet transform showed the best detection capability, compared with the case using optimal features selected from principal components.