Bias-Compensated Pseudo-measurement Tracking Filter Design in Line-of-Sight Coordinates

조민현,탁민제,박준현 한국항공우주학회 2021 International Journal of Aeronautical and Space Sc Vol.22 No.2

Target tracking using pseudo-measurement is widely used because of its simple implementation and computational advantages despite biased behavior. The biasedness of the pseudo-measurement tracking filter originates from a measurement-dependent Kalman gain and a state-dependent non-Gaussian biased noise. This paper proposes a bias-compensated pseudo-measurement filter considering a range rate measurement in the line-of-sight Cartesian coordinate system. Firstly, a pseudo-measurement model incorporating range rate measurements in line-of-sight coordinates is derived. The noise and measurement covariance of a proposed de-biased pseudo-measurement model are shown to be statistically consistent for highly noisy measurements. Secondly, a bias-compensated pseudo-measurement filter that adopts modified gain to suppress the estimation bias is formulated. The asymptotic stability of a proposed filter is further discussed. Lastly, simulation results show that a proposed de-biased pseudo-measurement filter is very effective in considering a range rate measurement.

품질이 관리된 스트레스 측정용 데이터셋 구축을 위한 제언

김태훈,나인섭 (사)한국스마트미디어학회 2024 스마트미디어저널 Vol.13 No.2

스트레스 측정용 데이터셋의 구축은 건강, 의료분야, 심리행동, 교육분야 등 현대의 다양한 응용 분야에서 핵심적인 역할을 수행하고 있다. 특히, 스트레스 측정용 인공지능 모델의 효율적인 훈련을 위해서는 다양한 편향성을 제거하고 품질 관리된 데이터셋을 구축하는 것이 중요하다. 본 논문에서는 다양한 편향성 제거를 통한 품질이 관리 된 스트레스 측정용 데이터셋 구축에 관하여 제안하였다. 이를 위해 스트레스 정의 및 측정도구 소개, 스트레스 인공지능 데이터 셋 구축과정, 품질향상을 위한 편향성 극복 전략 그리고 스트레스 데이터 수집시 고려사항을 제시하였다. 특히, 데이터셋 품질을 관리하기 위해 데이터셋 구축시 고려사항과, 발생할 수 있는 선택편향, 측정편향, 인과관계편향, 확증편향, 인공지능편향과 같은 다양한 편향성에 대해 검토하였다. 본 논문을 통해 스트레스 데이터 수집시 고려사항과 스트레스 데이터셋의 구축에서 발생할 수 있는 다양한 편향성을 체계적으로 이해하고, 이를 극복하여 품질이 보장된 데이터셋을 구축하는데 기여할 것으로 기대된다. The construction of a stress measurement dataset plays a crucial role in various modern applications. In particular, for the efficient training of artificial intelligence models for stress measurement, it is essential to compare various biases and construct a quality-controlled dataset. In this paper, we propose the construction of a stress measurement dataset with quality management through the comparison of various biases. To achieve this, we introduce stress definitions and measurement tools, the process of building an artificial intelligence stress dataset, strategies to overcome biases for quality improvement, and considerations for stress data collection. Specifically, to manage dataset quality, we discuss various biases such as selection bias, measurement bias, causal bias, confirmation bias, and artificial intelligence bias that may arise during stress data collection. Through this paper, we aim to systematically understand considerations for stress data collection and various biases that may occur during the construction of a stress dataset, contributing to the construction of a dataset with guaranteed quality by overcoming these biases.

GPS 기반 추적레이더 실시간 바이어스 추정 및 정보융합을 통한 발사체 추적 성능 개선 연구

송하룡,윤석영,한유수 한국항공우주학회 2015 한국항공우주학회 학술발표회 논문집 Vol.2015 No.4

다중 센서 추적시스템에서 측정치의 융합은 다수의 센서 데이터들을 좀더 정확한 하나의 트랙 정보를 갖도록 하기 위해 요구된다. 다중센서 시스템에서 센서 바이어스, 위치 불확실성과 안테나 지향에러와 같은 센서 등록 에러는 각각의 센서가 공통 좌표에 놓일 수 있도록 하기 위해 반드시 제거되어야 한다. 만약 센서등록 에러가 적절하게 보상되지 않는다면, 거대한 추적에러 또는 같은 목표물을 향한 다수의 허수 트랙이 발생하게 되어 추적에 실패하게 된다. 특히, 발사체 추적에 있어서 각각의 추적장비는 반드시 적절한 센서등록 과정을 거쳐야 하며, 이 후 다중센서 융합알고리즘을 활용하면 발사체 추적성능을 높이고 다중 추적 시스템에 정확한 지향입력으로 활용 가능하게 된다. 그러므로 본 논문에서는 실시간 바이어스 추정/제거 알고리즘과 다중 센서 융합 기법을 제안하였다. 제안된 바이어스 추정 알고리즘은 GPS 와 다중 레이더 간의 의사 바이어스 측정치를 활용하였고, 다중 센서 융합알고리즘 적용을 통해 추적 성능을 향상하였다. 시뮬레이션을 위하여, 실제 KSLV-I 3 차 발사 시 사용된 GPS 및 추적레이더 데이터를 사용하여 알고리즘의 성능을 테스트 하였다. Multisensor tracking scheme requires fusion of multiple observations to get a combined and more accurate target track. In multisensor system, sensor registration errors such as sensor bias, calibration error, site position uncertainty and antenna orientation must be corrected so that the individual sensor data are expressed in a common reference frame. If registration error is not properly compensated, the error can cause large tracking error or formation of multiple track on the same target. Especially for space launch vehicle (SLV) tracking system, each multiple observation lies on the same reference frame and then fused trajectory can be the best track for slaving data to aid other tracking systems. Hence, this paper describes an on-line bias estimation/correction and multi-sensor data fusion scheme for launch vehicle tracking system. The proposed bias estimation architecture is designed based on pseudo bias measurement which derived from error observation between GPS and radar measurements. Then, multiple sensor fusion scheme is adapted to enhance tracking performance. For simulation, real GPS and radar tracking data of the 3rd mission of KSLV-I is presented in experimental verification.

Investigating possible causes of bias in a progress test translation: an one-edged sword

Dario Cecilio-Fernandes,André Bremers,Carlos Fernando Collares,Wybe Nieuwland,Cees van der Vleuten,René A. Tio 한국의학교육학회 2019 Korean journal of medical education Vol.31 No.3

Purpose: Assessment in different languages should measure the same construct. However, item characteristics, such as item flaws and content, may favor one test-taker group over another. This is known as item bias. Although some studies have focused on item bias, little is known about item bias and its association with items characteristics. Therefore, this study investigated the association between item characteristics and bias. Methods: The University of Groningen offers both an international and a national bachelor’s program in medicine. Students in both programs take the same progress test, but the international progress test is literally translated into English from the Dutch version. Differential item functioning was calculated to analyze item bias in four subsequent progress tests. Items were also classified by their categories, number of alternatives, item flaw, item length, and whether it was a case-based question. Results: The proportion of items with bias ranged from 34% to 36% for the various tests. The number of items and the size of their bias was very similar in both programmes. We have identified that the more complex items with more alternatives favored the national students, whereas shorter items and fewer alternatives favored the international students. Conclusion: Although nearly 35% of all items contain bias, the distribution and the size of the bias were similar for both groups. The findings of this paper may be used to improve the writing process of the items, by avoiding some characteristics that may benefit one group whilst being a disadvantage for others.

GPS 기반 추적레이더 실시간 바이어스 추정 및 비동기 정보융합을 통한 발사체 추적 성능 개선

송하룡(Ha-Ryong Song) 한국산업정보학회 2015 한국산업정보학회논문지 Vol.20 No.6

다중센서 시스템에서 센서 바이어스를 제거하는 센서 등록 과정은 각각의 센서가 공통된 좌표를 갖게 하기 위해 반드시 필요하다. 만약 센서 등록 과정을 적절하게 처리하지 않는다면, 거대한 추적 에러 또는 같은 목표물을 향한 다수의 허수 트랙이 발생하게 되어 추적에 실패하게 된다. 특히, 발사체 추적에 있어서 각각의 추적 장비는 반드시 적절한 센서등록 과정을 거쳐야 하며, 이 후 다중센서 융합알고리즘을 활용하면 발사체 추적 성능을 높이고 다중 추적 시스템에 정확한 지향입력으로 활용 가능하게 된다. 본 논문에서는 실시간 바이어스 추정/제거 알고리즘과 비동기 다중 센서 융합 기법을 제안하였다. 제안된 바이어스 추정 알고리즘은 GPS와 다중 레이더 간의 의사 바이어스 측정치를 활용하였고, 비동기 센서 융합알고리즘 적용을 통해 추적 성능을 향상하였다. In the multi-sensor system, sensor registration errors such as a sensor bias must be corrected so that the individual sensor data are expressed in a common reference frame. If registration process is not properly executed, large tracking errors or formation of multiple track on the same target can be occured. Especially for launch vehicle tracking system, each multiple observation lies on the same reference frame and then fused trajectory can be the best track for slaving data. Hence, this paper describes an on-line bias estimation/correction and asynchronous sensor fusion for launch vehicle tracking. The bias estimation architecture is designed based on pseudo bias measurement which derived from error observation between GPS and radar measurements. Then, asynchronous sensor fusion is adapted to enhance tracking performance.

체계적 측정오차를 포함한 반응변수의 회귀모형 추론

송주원(Juwon Song) 한국자료분석학회 2021 Journal of the Korean Data Analysis Society Vol.23 No.6

회귀분석을 실시할 때 측정된 자료는 오차 없이 정확히 측정되었다고 가정하는 게 일반적이다. 하지만 실제 자료는 정확한 측정이 어려워 오차를 포함하여 측정되는 경우가 발생하며 이를 무시한 채 회귀분석을 실시한다면 회귀계수의 추정량에 편향이 발생할 수 있다. 회귀분석의 경우 설명변수가 오차를 포함하여 측정된 경우를 흔히 가정하지만 반응변수가 오차를 포함하여 측정된 경우에도 이를 보정하는 다양한 방법들이 제안되었다. 본 연구에서는 Nab et al.(2019)이 제안한 측정오차를 포함한 반응변수에 대한 선형 회귀분석에서 회귀계수의 보정 방법을 고려하였다. 특히 반응변수에 대한 체계적 측정오차 모형 하에서 선형 회귀분석의 회귀계수 보정 추정량의 특성을 살펴보았는데 이 추정량은 일치추정량이지만 불편추정량이 아니므로 유한표본에서는 편향이 발생할 수 있고 이에 영향을 미치는 요소들을 파악하기 위하여 모의실험을 실시하였다. 표본의 크기가 커짐에 따라 기울기 계수의 추정량에서 편향이 줄어들 뿐 아니라 분석 모형의 회귀계수값, 체계적 측정오차 모형의 회귀계수값, 분석모형의 결정계수, 체계적 측정오차 모형의 결정계수에 따라 편향 정도가 달라짐을 확인하였다. 또한 보정추정량에 대하여 델타방법을 사용하여 계산한 신뢰구간은 신뢰수준이 매우 높게 나타났다. Regression analysis assumes that variables are measured without errors. However, real data may include measurement errors due to various reasons, and it can cause bias in the estimation of regression parameters. Many researches have been conducted to adjust the bias of the regression parameters when the response variable is measured with errors. In this study, we consider the bias correction method suggested by Nab et al. (2019). Under the systematic measurement error model, it was shown that the bias corrected estimator of the linear regression parameters are consistent but biased. In this study, a simulation was conducted to evaluate performance of the suggested bias corrected estimator under the finite samples. It was found that the bias of the slope parameter was affected by the sample size, the size of the regression parameter in the analysis model, the size of the parameter in the systematic measurement error model, the coefficient of determination in the analysis model as well as in the systematic measurement error model. The coverage of the 95% confidence interval was very high when the delta method was applied for the bias corrected estimator.

Measurement Issues across Different Cultures

Lee, Ju-Hee,Jung, Duk-Yoo Korean Society of Nursing Science 2006 Journal of Korean Academy of Nursing Vol.36 No.8

Purpose. The purposes of this methodologic paper are to (1) describe theoretical background in conducting research across different cultures; (2) address measurement issues related to instrument administration; and (3) provide strategies to deal with measurement issues. Methods. A thorough review of the literature was conducted. A theoretical background is provided, and examples of administering instrument in studies are described. Results. When applying an instrument to different cultures, both equivalence and bias need to be established. Three levels of equivalence, i.e., construct equivalence, measurement unit equivalence, and full score comparability, need to be explained to maintain the same concept being measured. In this paper, sources of bias in construct, method, and item are discussed. Issues related to instrument administration in a cross-cultural study are described. Conclusion. Researchers need to acknowledge various group differences in concept and/or language that include a specific set of symbols and norms. There is a need to question the philosophical and conceptual appropriateness of an assessment measure that has been conceptualized and operationalized in a different culture. Additionally, testing different response formats such as narrowing response range can be considered to reduce bias.

Investigating the Effect of Training on Raters’ Bias toward Test Takers in Oral Proficiency Assessment : A FACETS Analysis

Houman Bijani,Mona Khabiri 아시아영어교육학회 2017 The Journal of Asia TEFL Vol.14 No.4

Typically, variability among raters in scoring and their bias is mediated through rater training. However, questions still remain about whether training can affect raters’ severity or leniency. Furthermore, few studies have looked at the differences between trained and untrained raters in oral assessment. Oral test scores of 200 test takers rated by 20 raters and were analyzed before and after a training program using the multifaceted Rasch measurement (MFRM). The results demonstrated the constructive impact of training programs in reducing raters’ biases and increasing their consistency measures. This study indicated that inexperienced raters benefited more from a training program than experienced raters and thus achieved higher measures of consistency afterward. It also demonstrated a higher biased interaction for test takers on the extreme ends of the oral ability continuum. The findings demonstrated that it is almost impossible to completely eradicate rater variability even through rater training. Therefore, rater training should be viewed as a procedure to establish within-rater consistency rather than between-rater consistency. Since this study showed that inexperienced raters can rate even more reliably than experienced ones after training, there is no evidence whereby decision makers can exclude inexperienced raters solely because of their lack of adequate experience. Consequently, decision makers need to use their budgets for establishing rater training programs for inexperienced raters instead.

A bias-correction for Cramér’s V and Tschuprow’s T

Wicher Bergsma 한국통계학회 2013 Journal of the Korean Statistical Society Vol.42 No.3

Cramér’s V and Tschuprow’s T are closely related nominal variable association measures,which are usually estimated by their empirical values. Although these estimators are consistent, they can have large bias for finite samples, making interpretation difficult. We propose a new and simple bias correction and show via simulations that, for larger than 2 × 2 tables, the newly obtained estimators outperform the classical (empirical) ones. For 2 × 2 tables performance is comparable. The larger the table and the smaller the sample size, the greater the superiority of the new estimators.

적응추정기를 이용한 강의 수질 감시 시스템 설계

김창일,권성숙,김경식,김경연 濟州大學校 情報通信硏究所 1998 情報通信硏究所論文集 Vol.1 No.-

An adaptive estimator to estimate the water quality in a river under unknown biased measurements environment is developed. In modeling the unknown measurement bias, it is assumed that the bias sequence is governed by semi-Markov process. By incorporating the semi-Markov probability concepts into the Bayesian estimation theory. an effective adaptive estimator which consists of parallel Kalman-type filters is obtained. Computer simulation results for the multiple-reach river system show that the proposed adaptive estimator have good estimation performance in spite of the unknown randomly switching measurement bias.