http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
- 中文 을 입력하시려면 zhongwen을 입력하시고 space를누르시면됩니다.
- 北京 을 입력하시려면 beijing을 입력하시고 space를 누르시면 됩니다.
RISS 인기검색어
- 검색키워드
- 저자 : 양길모 (검색결과 44 건)
- 무료
- 기관 내 무료
- 유료
-
The Effects on the Top-N Accuracy of Response Characteristics
양길모,주상열,김선옥 한국자료분석학회 2010 Journal of the Korean Data Analysis Society Vol.12 No.1
This research is the study of how response characteristics of users affect the accuracy of Top-N. Also, this study in Top-N recommendation is the base research about the determination of N in Top-N. According to the response characteristics(response mean of each individual user, standard deviation, skewness, and kurtosis), the difference of the accuracy of Top-N is analyzed. As a result of this experiment, the accuracy of Top-N of users can be classified by response mean of each individual user, standard deviation, skewness, and kurtosis. The accuracy of Top-N in e-commerce is a tool that can give trust to customers who visit the e-commercial sites. This is also expected to be used on the expanding m-commerce market. The recommender system in on-line business transaction is a useful tool, but making the lists of 'Top-N is not a simple matter. For a further study, formula to determine the N of Top-N should be extracted.
-
메모리 기반 협력적 필터링 알고리즘의 선호도 예측 특성에 관한 연구
양길모,이희춘,박영서 한국자료분석학회 2008 Journal of the Korean Data Analysis Society Vol.10 No.4
This study describes customer preference prediction of memory-based collaborative filtering algorithm about 100K and 1million MovieLens dataset. We found that prediction accuracy of Correspondence Mean Algorithm is more stable than Neighborhood-based Collaborative Filtering Algorithm by comparing real with prediction preference through NBCFA and CMA in customer preference prediction, then knew that the superiority of CMA by comparing mutual relation between real and prediction preference evaluation. Also, the prediction accuracy about the frequency between 1 and 2 of these algorithm decreases, and according to the analysis result of this study providing the basic materials for post-study is the goal of this study to improve prediction accuracy. 본 연구는 100K와 1million MovieLens dataset에 대한 메모리 기반 협력적 필터링 알고리즘의 고객 선호도 예측 특성에 대하여 연구하였다. 이웃 기반의 협력적 필터링 알고리즘과 대응 평균 알고리즘의 실제 선호도에 대한 예측 선호도의 특성을 실제 선호도 평가에 따라 비교함으로써 고객 선호도 예측에 있어 대응평균 알고리즘의 예측 정확도가 이웃 기반의 협력적 필터링 알고리즘에 비하여 안정적임을 알 수 있었으며 실제 선호도 평가치와 예측 선호도 평가치의 상관관계를 비교함으로써 대응평균 알고리즘의 우수성을 알 수 있었다. 또한 두 알고리즘 모두 선호도 평가치의 빈도가 작은 1과 2에서의 예측 정확도가 떨어짐을 알 수 있었으며 본 연구의 분석결과를 토대로 예측 정확도를 높이기 위한 차기 연구의 기초 자료를 제공하는 것을 본 연구의 목적으로 한다.
-
양길모,고학균,홍지향 한국농업기계학회 2003 바이오시스템공학 Vol.28 No.2
The study was conducted to develop precision control system for cold storage facility that could offer safe storage environment for green grocery. For that reason, neuro-fuzzy control system with learning ability algorithm and single chip neuro-fuzzy micro controller was developed for cold storage facility. Dynamic characteristics and hunting of neuro-fuzzy control system were far superior to on-off and fuzzy control system. Dynamic characteristics of temperature were faster than on-off control system by 1,555 seconds(123% faster) and fuzzy control system by 460 seconds(36.4% faster). When system, and ±0.4˚C in fuzzy control system, and ±0.3˚C in neuro-fuzzy control system. Hunting of humidity and wind velocity was also controlled precisely by 70 to 72.5% and 1m/s For storage experiment with onion, characteristics of neuro-fuzzy control system were tested. Dynamic characteristics of neuro-fuzzy control system made cold storage facility conducted precooling ability and minimized hunting.
-
-
양길모,최규홍,조남홍,박종률 한국농업기계학회 2005 바이오시스템공학 Vol.30 No.3
Grading and sorting an indeterminate form of agricultural products such as sweet potatoes and potatoes are a labor intensive job because its shape and size are various and complicate. It costs a great deal to sort sweet potato in an indeterminate forms. There is a great need for an automatic grader for the potatoes. Machine vision is the promising solution for this purpose. The optical indices for qualifying weight and appearance quality such as shape, color, defects, etc. were obtained and an on-line sorting system was developed. The results are summarized as follows. Sorting system combined with an on-line inspection device was composed of 5 sections, human inspection, feeding, illumination chamber, image processing & control, and grading & discharging. The algorithms to compute geometrical parameters related to the external guality were developed and implemented for sorting the deformed sweet potatoes. Grading accuracy by image processing was 96.4% and the processing capacity was 10,800 pieces per hour. 본 연구에서는 인력에 의존하고 있는 부정형 농산물의 선별작업을 기계화할 수 있도록 영상처리를 이용한 자동선별시스템을 개발하고 고구마를 대상으로 하여 그 성능을 평가하였다. 시작기는 무게, 형상, 휨, 손상, 색택을 영상처리를 이용하여 정밀하게 선별할 수 있었다. 구체적인 연구결과를 요약하면 다음과 같다.등급판정 인자는 각 산지에서 생산되는 서로 다른 품종의 고구마를 각 등급별로 구입을 하여 길이, 장경, 단경, 무게를 측정한 후 기하학적 특성을 분석하여 구명하였다.선별용 컵은 이들 기하학적 특성에 기초하여 가로×세로×높이가 240×128×42 mm로 하였고, 하부낙하 슬라이드 방식을 적용하였다.시작기는 길이 9,663 mm, 폭 1,006 mm, 높이 799 mm의 크기를 가지고 있으며, 구성은 육안선별부, 공급부, 영상획득부, 무게측정부, 영상처리 및 제어부, 등급별 배출부, 자동 상자담기로 일괄처리할 수 있도록 고안하였다. 특히, 자동공급부는 브러쉬 롤러와 횡방향 공급, 종방향 정렬방법으로 낱개공급 장치를 개발하였다.영상처리 시스템은 빛을 조사할 수 있는 챔버, 영상을 획득하는 CCD 카메라, 시료의 전체 영상을 획득할 수 있도록 도와주는 회전형 고무롤 컨베이어로 구성하였으며, 시료당 3번의 연속 촬영에 의해 전체영상을 획득하여 등급판정을 하였다. 무게측정은 로드셀을 이용하였고 형상은 길이/장경비로, 휨은 굽은 정도, 손상은 화소값, 색택은 영상처리의 문턱값을 적용하였다. 등급판정은 이들 인자들을 조합하여 판정하였다.고구마 영상처리 시스템의 작업성능은 시간당 최대 10,800개를 선별할 수 있으며, 선별정밀도는 96.4%로 양호한 성능을 보였다. 이때 작업 중 손상은 없는 것으로 나타났다.
-
3차원 CFD 시뮬레이션을 이용한 돔형 저온저장고 개발
양길모,고학균,홍지향 한국농업기계학회 2003 바이오시스템공학 Vol.28 No.1
This study was conducted to develop proper model for cold storage facility that could offer uniform heat movement and air movement for green grocery and improve improper of the existing container-type cold storge facility.For that reason, new model(dome-type) cold storage facility was developed using 3-DCFD(computational fluid dynamics) simulation. The size was 6m×6m×5m. Its size and configuration were same to simulation model. Unit cooler was designed to send cold air in 4 side ways. A dome-type cold storage facility showed uniform distributions of air temperature and velocity because cold air was forced to move down along the ceiling and the wall and then circulated to the unit cooler from the central part of the floor. Dome-type cold storage facility also showed by low wind velocity, below 1 m/s that could minimized cold damage and quality deterioration.
-
-
양길모,고학균,조성인 한국농업기계학회 2000 바이오시스템공학 Vol.25 No.2
Low temperature storage method is to increase the value of agricultural products by reducing quality loss and regulate consignment time by controlling respiration rates of agricultural products. Respiration rate of agricultural products depends on several factors such as temperature, moisture, gas composition and a microbe inside the storage room. Temperature is the most important factor among these, which affects respiration rate and causes low or high temperature damage. Fuzzy logic was used to control the temperature of a storage room ,which uses information of uncertain facts and mathematical model for room temperature control . Room temperature was controlled better by using fuzzy logic control method rather than on-off control method. Refrigerant flow rates and temperature deviations were measured for on-off system using TEV(temperature expansion valve) and for fuzzy system using EEV(Electrical Expansion Valve) . Temperature of the Storage room was lowered faster by using fuzzy system than on -off system. Temperature deviation was -0.6~+0.9$^{\circ}C$ for on-off system and $\pm$0.2$^{\circ}C$ for fuzzy system developed. Temperature deviation and variation of temperature deviation were used as inout parameters for fuzzy system. The most suitable input and output value were found by experiment. Cooling rate of the storage room decreased while temperature deviation increased for the sampling time of 20 sec.
-
-
KCI
스콜라연관 검색어 추천
이 검색어로 많이 본 자료
활용도 높은 자료
