• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2006년도 Proceedings of ISRS 2006 PORSEC Volume II
• /
• pp.635-638
• /
• 2006

Regression of conventional prediction techniques in data mining uses the model which is generated from the training step. This model is applied to new input data without any change. If this model is applied directly to time series, the rate of prediction accuracy will be decreased. This paper proposes an incremental regression for time series prediction like typhoon track prediction. This technique considers the characteristic of time series which may be changed over time. It is composed of two steps. The first step executes a fractional process for applying input data to the regression model. The second step updates the model by using its information as new data. Additionally, the model is maintained by only recent data in a queue. This approach has the following two advantages. It maintains the minimum information of the model by using a matrix, so space complexity is reduced. Moreover, it prevents the increment of error rate by updating the model over time. Accuracy rate of the proposed method is measured by RME(Relative Mean Error) and RMSE(Root Mean Square Error). The results of typhoon track prediction experiment are performed by the proposed technique IMLR(Incremental Multiple Linear Regression) is more efficient than those of MLR(Multiple Linear Regression) and SVR(Support Vector Regression).

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제13권3호
• /
• pp.178-185
• /
• 2013

As computer technology continues to develop, computer networks are now widely used. As a result, there are many new intrusion types appearing and information security is becoming increasingly important. Although there are many kinds of intrusion detection systems deployed to protect our modern networks, we are constantly hearing reports of hackers causing major disruptions. Since existing technologies all have some disadvantages, we utilize algorithms, such as the fuzzy C-means (FCM) and the support vector machine (SVM) algorithms to improve these technologies. Using these two algorithms alone has some disadvantages leading to a low classification accuracy rate. In the case of FCM, self-adaptability is weak, and the algorithm is sensitive to the initial value, vulnerable to the impact of noise and isolated points, and can easily converge to local extrema among other defects. These weaknesses may yield an unsatisfactory detection result with a low detection rate. We use a genetic algorithm (GA) to help resolve these problems. Our experimental results show that the combined GA and FCM algorithm's accuracy rate is approximately 30% higher than that of the standard FCM thereby demonstrating that our approach is substantially more effective.

• 한국가축번식학회지
• /
• 제13권3호
• /
• pp.149-156
• /
• 1989

These experiment was carried out to improve the precision of early pregnancy diagnosis in dairy cattle. Changes in progesterone concentration of milk were measured by Enzyme Immunoassay(EIA) in 73 cows up to 21 days after insemination. The average concentraton of progesterone in milk was 1.9ng/ml at eatrus ; it increased to 17.8ng/ml on day 14, and thereafter it declined to 4.3ng/ml on day 21 in nonpregnant cows. Whereas in pregnant animals, it was maintained and elevated further to 22.2ng/ml on day 21. The accuracy of the pregnancy diagnosis for animals classified as positive (pregnant ; over 10ng/ml and decreasing rate<1.5) and negative (non-pregnant ; under 7ng/ml and decreasing rate>1.5) were 95% and 100% respectively. The samples ranging between 7ng/ml and 10ng/ml were classified as positive (decreasing rate<1.5) and negative(>1.5), which accuracy was 54.5% and 100% respectively. However this range appears to be of the most interest for the veterinary practitioner since cows in proestrus or early interestrus tend to have milk progesterone levels within this value. Causes for the insufficient precision of false pregnancy diagnosis are discussed.

• 한국콘텐츠학회논문지
• /
• 제14권9호
• /
• pp.476-483
• /
• 2014

본 논문은 국내 표정 연구에 적합한 얼굴 표정 이미지를 제작하는 것에 목적을 두었다. 이를 위해서 1980년대 태생의 한국인의 표준 형상에 FACS-Action Unit을 결합하여, KSFI(Korean Standard Facial Image) AU set를 제작하였다. KSFI의 객관성을 확보하기 위해 6가지 기본 감성(슬픔, 행복, 혐오, 공포, 화남, 놀람) 이미지를 제작하여, 감성 별 인식 정확률과 얼굴 요소의 감성인식 기여도를 평가하였다. 실험 결과, 정확률이 높은 행복, 놀람, 슬픔, 분노의 이미지의 경우 주로 눈과 입의 얼굴 요소를 통해 감성을 판단하였다. 이러한 연구 결과를 통해 본 연구에서는 표정 이미지의 AU 변경할 수 있는 KSFI 콘텐츠를 제안하였다. 향후 KSFI가 감성 인식률 향상에 기여할 수 있는 학습 콘텐츠로서의 역할을 할 수 있을 것으로 사료된다.

• 한국자동차공학회논문집
• /
• 제12권6호
• /
• pp.182-190
• /
• 2004

This paper presents reconstruction analysis of vehicle trajectory using records of a developed black box, and results of validation tests. For reconstruction of vehicle trajectory, the black box records the longitudinal and lateral accelerations and yaw-rate of vehicle during a pre-defined time period before and after the accident. One 2-axis accelerometer is used for measuring accelerations, and one vibrating structure type gyroscope is used for measuring yaw-rate of vehicle. The vehicle's planar trajectory can be reconstructed by integrating twice accelerations along longitudinal and lateral directions with yaw-rate values. However, there may be many kinds of errors in sensor measurements. The causes of errors are as follows: mis-alignment, low frequency offset drift, high frequency noise, and projecting 3-dimensional motion into 2-dimensional motion. Therefore, some procedures are taken for error compensation. In order to evaluate the reliability and the accuracy of trajectory reconstruction results, the black box was mounted on a passenger car. The vehicle was driven and tested along various specified lanes. Through the tests, the accuracy and usefulness of the reconstruction analysis have been validated.

• 한국기후변화학회지
• /
• 제5권4호
• /
• pp.323-329
• /
• 2014

Since the terrain of Korea is complex, micro- as well as meso-climate variability is extreme by locations in Korea. In particular, air temperature of agricultural fields is influenced by topographic features of the surroundings making accurate interpolation of regional meteorological data from point-measured data. This study was carried out to compare spatial interpolation methods to estimate air temperature in agricultural fields surrounded by rugged terrains in South Korea. Four spatial interpolation methods including Inverse Distance Weighting (IDW), Spline, Ordinary Kriging (with the temperature lapse rate) and Cokriging were tested to estimate monthly air temperature of unobserved stations. Monthly measured data sets (minimum and maximum air temperature) from 588 automatic weather system(AWS) locations in South Korea were used to generate the gridded air temperature surface. As the result, temperature lapse rate improved accuracy of all of interpolation methods, especially, spline showed the lowest RMSE of spatial interpolation methods in both maximum and minimum air temperature estimation.

• Nuclear Engineering and Technology
• /
• 제54권9호
• /
• pp.3398-3402
• /
• 2022

We fabricated a 15 m long position-sensitive plastic scintillation optical fiber (PSOF) detector consisting of a PSOF, two photomultiplier tubes, four fast amplifiers, and a digitizer. A single PSOF was used as a sensing part to estimate the gamma-ray source position, and ¹³⁷Cs, an uncollimated solid-disk-type radioactive isotope, was used as a gamma-ray emitter. To improve the sensitivity, accuracy, and measurement time of a PSOF detector compared to those of previous studies, the performance of the amplifier was optimized, and the digital signal processing (DSP) was newly designed in this study. Moreover, we could measure very low dose rates of gamma-rays with high sensitivity and accuracy in a very short time using our proposed PSOF detector. The results of this study indicate that it is possible to accurately and quickly locate the position of a very low dose rate gamma-ray source in a wide range of contaminated areas using the proposed position-sensitive PSOF detector.

• International Journal of Computer Science & Network Security
• /
• 제22권3호
• /
• pp.61-66
• /
• 2022

Object classification is one of the main fields in neural networks and has attracted the interest of many researchers. Although there have been vast advancements in this area, still there are many challenges that are faced even in the current era due to its inefficiency in handling large data, linguistic and dimensional complexities. Powerful hardware and software approaches in Neural Networks such as Deep Neural Networks present efficient mechanisms and contribute a lot to the field of object recognition as well as to handle time series classification. Due to the high rate of accuracy in terms of prediction rate, a neural network is often preferred in applications that require identification, segmentation, and detection based on features. Neural networks self-learning ability has revolutionized computing power and has its application in numerous fields such as powering unmanned self-driving vehicles, speech recognition, etc. In this paper, the experiment is conducted to implement a neural approach to identify numbers in different formats without human intervention. Measures are taken to improve the efficiency of the machines to classify and identify numbers. Experimental results show the importance of having training sets to achieve better recognition accuracy.

• 한국학교보건학회지
• /
• 제27권3호
• /
• pp.130-139
• /
• 2014

Purpose: The object of this study was to compare the educational effect about self-efficacy and the quality of chest compressions of Hands-only CPR and Basic CPR. Methods: It's a nonequivalent control group pre-post repeated quasi-experiment study conducted with entire fifth grade students belong to one school in H city. The study participants are 68 persons, and data were collected from December 2, 2013 to February 7, 2014. Self-efficacy was measured by 10 items, and the quality of chest compressions was measured by 5 variables which are average compression depth(mm), average rate (n/min), average count per minutes (n), abnormal placement (n), compression accuracy (%). Results: Self-efficacy of the experimental group and control group showed no significant difference but showed significant difference over time and was the highest at posttest 1 (immediately after education), the lowest at pretest (before education), middle at posttest 2 (8weeks after education) (p<.001). Experimental group was significantly higher than control group in average rate per minute. At posttest 1, experimental group was $130.0{\pm}9.38$ times, control group was $95.1{\pm}11.82$ times. At posttest2, experimental group was $124.0{\pm}14.89$ times, control group was $90.8{\pm}14.89$ times.(p<.001). Average rate (n/min) was significantly declined at control group in the quality of chest compressions over time (t=-2.400, p=.022). Average count per minute and compression accuracy were declined significantly so it were not maintained to posttest2. Conclusion: We need continuous CPR education because self-efficacy of CPR getting lower significantly over time. Hands-only CPR can't be seen as a way to increase the CPR ability of elementary school students having difficulty to perform artificial breathing. And, because the effect of education is not maintained 8wks after training, the technique centered repeated training is needed and a method which can increase compression accuracy is also needed.

• 한국항행학회논문지
• /
• 제13권4호
• /
• pp.482-489
• /
• 2009

이 논문에서는 AWGN환경에서 발생하는 가우시안 잡음과 발진기에서 생기는 시스템 잡음을 제거할 수 있는 ALE(Adaptive Line Enhancer) 기반의 새로운 적응형 잡음 제거기를 이용한 CW(Continuous-Wave) 바이오 레이더를 제안한다. 최근에 CW 바이오 레이더를 이용해서 심장박동과 호흡을 검출하는 연구가 여러 연구기관에서 진행 되고 있다. 그러나 이 연구들은 기존 CW 바이오 레이더가 가우시안 잡음에 취약하고 그로 인해 심장박동 검출정확도도 떨어진다는 점을 설명을 하고 있긴 하지만, 그 잡음을 효과적으로 없앨 수 있는 방안은 계속 연구 중에 있다. 본 논문에서는 기저대역 신호에 포함된 잡음을 효과적으로 제거할 수 있는 ALE기반의 적응형 잡음 제거기를 적용한 것을 제안한다. 또한 타겟의 위치에 따른 복조의 민감함에 강점을 가진 quadrature 수신기를 통과한 잡음이 포함된 기저대역 신호에서 잡음만을 효과적으로 제거함으로 인해 심장박동 검출 정확도를 향상시키는 것을 모의실험을 통해 비교 분석해 본다.