http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
권태희,장재원,최선우,성기정,Kwon Tae-Hee,Chang Jae-Won,Choi Sun-Woo,Seong Kie-Jeong 한국군사과학기술학회 2005 한국군사과학기술학회지 Vol.8 No.1
After the development of the Firefly, flight tests have been performed to verify the performance and get the parameters for the mathematical model of the aircraft. The flight test data is used to get parameters for the mathematical model of the aircraft through the parameter identification process. An arbitrary control input is applied to the test flight which is a part of parameter identification process. A square wave has been used a control input which is called Doublet signal. The aspect of the signal is same length and magnitude in both (+) and (-) directions such as sine wave. The Doublet signal is composed of a dominant frequency and many high frequencies, so that it is appropriate signal to excite the motion of an aircraft. In this paper, the control input of the flight test data has been analyzed to check the efficiency of the control input using DFT(Discrete Fourier Transform). From the result of analysis, an alternative input was extracted.
생명보험(生命保險) 유진사(有診査) 및 무진사가입자(無診査加入者)의 사망(死亡)에 관한 고찰(考察)
권태희,Kwon, Tae-Hee 한국생명보험의학회 1984 保險醫學會誌 Vol.1 No.1
In Korea, life insurance policies are sold to the policy holders by insuring either the insureds undergo a medical examination at a clinic or the insureds' report their history of diseases ever experienced that replaces the medical examination. This study aimed to measure the level of death rates for the insureds between those who received medical examination and those who did not receive medical examination, and to examine differences of the rates in terms of the insureds' characteristics such as age, sex, cause of death and duration. A total number of 32,358 insureds were selected for the population of this study from the D. Life Insurance Company located in Seoul City. Out of the 32,358 insureds, 2,997 received medical examination and the rest of 29,381 did not received any medical examination. Results of analysis are summarized as follows: 1. Death rate per 100,000 insureds for the all was 19.3 in the first year, 96.3 in the second year, 143.8 in the third year 93.4 in the fourth year. For the group of medical examination received, the rate was zero in the first year, 41.3 in the second year, 55.4 in the third year and 268.8 in the fourth year, and for the group of non-medically examined the rate was 21.3 in the first year, 101.9 in the second year, 152.2 in the third year and 76.8 in the fourth year. The levels of death rates between the insureds with medical examsination and the inureds without medical examination were non-significant in the differences by duration except the levels of the third year, which indicated the death rate of non-medically examined group was higher than that of the medically examined group. 2. 73.0 per cent of the total deaths observed during the insured period were caused by various diseases and the rest of 27.0 per cent deaths were due to accidents. For the group of medical examination received, 55.6 per cent deaths were caused by diseases, and for the group of nonmedically examined, 74.7 per cent of deaths were due to diseases. 3. cancer was the most frequent cause which accounted for 22.0 per cent of the total deaths. Proportion of deaths due to cancer from the group of medical examination received was 22.2 per cent, and the corresponding rate for the group of non-medically examined also showed high rate of 22.0 per cent.
데이터 분석 도구 성능 비교 연구 -기계 학습을 적용하여-
권태희 ( Tae-hee Kwon ) 한국정보처리학회 2016 한국정보처리학회 학술대회논문집 Vol.23 No.2
빅데이터 시대가 도래되면서 과거와 비교할 수 없을 만큼의 방대하고 다양한 데이터가 생산됨에 따라 기존의 데이터 분석 도구의 사용은 한계에 부딪히게 되었다. 따라서 기존의 분석 도구보다 효율적이고 정확성이 높은 데이터 분석 도구를 필요로 하게 되었고, 빅데이터를 처리할 수 있는 분석 도구들에 대한 많은 연구들이 진행되어 왔다. R과 Apache Spark는 대표적인 데이터 분석 도구로 기계 학습을 위한 기능을 제공하고 있다. 본 논문에서는 기계 학습을 활용하여 두 개의 널리 알려진 데이터 분석 도구인 R과 Apache Spark의 데이터 분석 성능을 비교함으로써 보다 효율적이고 정확성이 높은 도구를 모색하고자 한다.