Comparative Study between Rate Parity and Rate Disparity Hotels : 호텔 객실요금 정책의 일관성에 대한 비교 연구

김태연 세종대학교 관광대학원 2013 국내박사

The purpose of this study is to identify the effectiveness of rate parity of hotel online pricing. The effectiveness of rate parity is evaluated by comparing 2 different groups of hotels, one which maintains rate parity and another which breaks rate parity. This evaluation is further carried out to examine from 3 different research questions, which are financial, operational, and managerial ineffectiveness. The data of this paper is based on the actual 3-year secondary time series, which are compared on monthly basis. Data were collected from four out of eight 5-star luxury business city hotels located in Gangnam Business District of Seoul, Korea in order to honor homogeneity, and external variables in consideration are similar facilities, location, star ratings, guest type, total number of rooms, average daily rate and primary target market. Two hotels in our sample were offering lower non-qualified selling rates to various local distribution channels for more than 3 consecutive years from 2005 especially to local On-Line Travel Agencies (OTA) and this group of hotels is named ‘rate parity breaking hotels’. Two other hotels in our sample were offering the same selling rates throughout all distribution channels and this group of hotels is named ‘rate parity pursuing hotels’. The extensive investigation of the time series data exhibited that there was not much difference in financial outcomes of the 2 groups of hotels. However, ADR decreased and customer type changed from business travelers to local leisure travelers for rate parity breaking hotels. Maintaining revenue while ADR is decreasing indicates that rate parity breaking hotels filled more rooms with lower rates, which in turn would have required more labor of front line employees. This burnout situation was also discussed by interviewing actual hotel front line employees and managers and found 3 stages of burnout. KEYWORDS: Rate Parity, Dynamic Pricing Strategy, Online Pricing, Revenue Management, and Distribution.

KOSPI 200 주가지수 옵션의 Put-Call Parity 이론에 대한 실증연구

김종호 충남대학교 2001 국내석사

Existing empirical studies of the put-call parity condition say frequent, substantial violations. An important problem in interpreting these results is that these studies all investigate American options. While some of these studies attempt to reduce the effect of possible early exercise on their tests, they cannot fully account for the effect of early exercise. Therefore, it is not possible to conslude from these studies whether, or to what extent, observed put-call parity violations are due to market inefficienty or due to the value of early exercise. This study avoid the early exercise problem by testing put-call parity using European options. This study finds violations that are much less frequent and smaller than the studies using American options. Moreever, these viloations may reflect premia for liquidity(immediacy) risk. Some forms fo contigent claim valuation are based on the no-arvitrage principle that uses close substitutes to price financial assets and contracts. The appeal of this principle is that is only requires a few rational agents who stand ready to eliminate any arbitrage opportunity. The no-arbitrage principle, therefore, can provide a simple way to test market dfficiency without imposing restrictions on agent's preferences and the stochastic disgrbutions governing asset returns. The put-call parity condition formalized by Stoll(1969) uses the no-arbitrage principle to price put (call) optionns relative to call (put) options. This study begins by reviewing the put-call parity formalized by Stoll(1969), but allowing for transaction costs, bid-ask spreads, and dividends. Define: P_b and P_a = current bid and ask prices of a European put option on one S&p share expiring on dat T with an execise price of X C_b and C_a = current bid and ask prices of an identical call option S_b and S_a = current bid and aks prices of one S&P share from today to date T D = present value of the dividend flow per S&P share from today to date T t_i = transactions costs of buying or selling asset I (i = p,c,s for puts, calls, and shares) λ_s = percent of proceeds available to S&P short-sellers γ_B and γ_L = (direct) riskless borrowing and lending interest rates from today to dat T. Two trading strategies define put-call parity. The first is to buy a portpolio of one S&P share, buy one put, and write one call. The strategy always yields $X on expiration date and its cost today of (S_a-D+t_s)+(P_a+t_p)-(C_b-T_c) is borrowed at γ_B. The second strategy is the reverse of the first. The investor much always pay $X on expiration day, and the proceeds today of (λ_s(S_b-D)-t_s)+(P_b-t_p)-(C_a+t_c) are invested at γ_L. Put-Call parity implies (1) X - [(S_a-D+t_s)+(P_a+t_p)-(C_b-t_c)}(1+γ_B)≤0 (2) [(λ_s(S_b-D)-t_s)+(P_b-t_p)-(C_a+t_c)](1+γ_L)-X≤0 Define R_L and R_B, the synthetic riskless lending and borrowing rates, as (3) R_L≡X/(S_a-D+t_s)+(P_a+t_p)-(C_b-t_c) - 1 (4) R_B≡X/(λ_s(S_b-D)-t_s)+(P_b-t_p)-(C_a+t_c) - 1 (5) γ_B-R_L ≥0, (6) R_B-γ_L≥0. This study tests the put-call parity condition using KOSPI 200 index option. It allows the disentanglesment of the issue of market efficiency from the issue of an early execrise premium. This paper documents some deviations from put-call parity during 1999, and these deviations show some differeces between first and second half of the year. More importantly, the diviations less frequent and smaller than these found in earlier put-call parity studies using American options. Furthermore, the results may suggest that the trading strategies underlying put-call parity are subject to significant liquidity risk. Variations in the deviations from put-call parity bounds may be systematically related to the liquidity risk in the stock and option markets.

나노구조의 광학적 특성 및 이미징 방법에 대한 연구

최지훈 국민대학교 일반대학원 2022 국내박사

We design and simulate the on-chip nanostructure for optical circuits. We demonstrate the single resonator, broadband resonator, and single port coherent perfect absorption in silicon-on-insulator and propose the reverse circulation mode using a microring resonator with Parity-Time structure. Additionally, we propose the imaging method for vascular patterns in the dermis and design and investigate experimentally the linear Fresnel lens. Time-reversed lasing, or coherent perfect absorption (CPA) has great potential for application to optical circuits such as sensors, optical switches, modulators, and logic gates because light at a specific wavelength can be perfectly absorbed by using resonance. Typically, CPA uses a single resonator with high quality (Q) factor. As a result, absorption spectra have a very narrow linewidth. Therefore, CPA with a single resonator is sensitive to the change of environment such as temperature. Narrow bandwidth can be greatly utilized for sensors with high sensitivity but can be tricky to use at on-chip optical circuits. Perfect absorption spectrum width in CPA can be broadened by using a multi-cavity structure. The basic principle is as follows. At the resonance, the phase of the wave at the boundary is zero. At telecommunication wavelength, 1550 nm, silicon, and silicon dioxide have no optical absorption but CPA can occur not only with optical absorption but also with all forms of other losses such as scattering and coupling. To emphasize the perfect extinction of light inside the cavity due to loss, we name it coherent perfect loss (CPL). We can achieve the conditions of CPL using resonators with scatterers. Also, the CPL mode with even symmetry can be reduced to a single-port CPL when a perfect magnetic conductor (PMC) boundary is applied. The boundary which acts like a PMC boundary, here known as a PMC-like boundary, can be realized by adjusting the phase shift of the reflection from the PC when the wavelength of the light is within the photonic bandgap wavelength range. We demonstrate on-chip CPL with a single resonator and broadband resonator using PC nanobeam in 3-dimensional FEM simulation (COMSOL Multiphysics). In CPL with a single resonator, we can find the perfect loss at a wavelength of 1550 nm with the full width at half max (FWHM) of 1.76 nm. The broadband CPL can be achieved in photonic crystal (PC) nanobeam using different hole sizes. If the hole size of the PC is modified, the effective refractive index will also be changed. Therefore, multi-cavities are achieved in a PC. The obtained perfect loss region with a PC nanobeam with broadband CPL is between 1549 nm and 1551 nm. We confirm that the spectral width of broadband CPL is certainly broader than that of CPL with a single resonator. Our designed CPL can be simply fabricated because it consists only of dielectric materials. The proposed CPL, therefore, has a cost-effective fabrication procedure. We numerically demonstrated single-port CPL with a Fabry–Perot resonator in a PC nanobeam by using a PMC-like boundary. We designed and optimized a simple Fabry–Perot resonator and coupler in nanobeam to get the PMC-like boundary. To satisfy the loss condition in CPL, we controlled the coupling loss in the resonator by modifying the lattice constant of the PC used for coupling. By optimizing the coupling loss, we achieved zero reflection (CPL) in a single port with a PMC-like boundary. Using the numerical analysis and FEM, we demonstrated the reversed circulation mode (RCM) in a single microring resonator (MRR) with Parity-Time (PT)-structure. For RCM mode, the direction of outgoing light is reversed from the direction of the incident light, unlike typical MRR. The RCM is unique in the broad-spectrum range. We find the origin of the RCM at PT-MRR coupled with waveguide (2-ports) by numerical analysis with directional coupling analysis using the transfer matrix method. We define the RCM at PT-MRR numerically by calculating the contrast between reflection and transmission at the waveguide for coupling. In FEM simulation, we designed PT-MRR coupled with 2 waveguides (4-ports) and demonstrated RCM. RCM at PT-MRR has great potential such as optical logic gates, amplifying mirrors, optical cross connectors, and optical filters in optical circuits. Because vascular patterns inside skin tissue are unique to individuals and impossible to counterfeit, they can be used for biometric identification. However, due to light scattering by skin tissue, obtaining a clear image without imaging the transmitted light using a lens is difficult, which in turn results in a large detection system. In this study, we present a novel method for imaging vascular patterns that use image scanning with reflected light and no lenses. We numerically confirm that the image scanning method produces a clearer image than imaging methods that do not employ the scanning method. In addition, using a smaller light source and detector produces a clearer image of vascular patterns. At a scan pixel size of 2 μm, we determined that the degree of cross-correlation was G=0.74. We designed a linear Fresnel lens (LFL) for harvesting solar energy. The optimized LFL was fabricated with dimensions of 100 (W)×100 (H)×5 (T) mm^3, a focal length of 100 mm, and a fixed pitch size of 500 nm. The maximum focal efficiency is approximately 91%, and the half width at half maximum (HWHM) value of the focus size is approximately 130 μm. We also investigated the robustness of the LFL to changes in the incident angle, draft angle, and detection position. Simulations revealed a focal efficiency above 85% with an incident angle up to 1°, a draft angle up to 10°, and a detection position up to ±10 mm. We fabricated the LFL by using an injection molding process. The fabricated lens had a maximum focal efficiency of approximately 88% and a full width at half max (FWHM) of about 2.5 mm. Finally, we designed the mode decomposer using the inverse design on SOI. For inverse design, we can optimize structure along electric/magnetic fields. The inverse design is studied in various ways such as spectrometer, mode converter, and mode decomposer for compact devices. The mode decomposer consists of 7.5 μm of width of the waveguide for input and a 12-channel output waveguide with 500 nm of the width. The design region has a size of 10×10 μm^2. The inverse design is performed in a 2-dimension using optimization effective refractive index. 12-channel mode decomposer is evaluated in a 3-dimensional finite-difference time-domain simulation. 12-channel mode decomposer has transmission efficiency up to 39% and mode crosstalk below 11%. We adopted the scattering region to mix mode at waveguide with a different ratio along wavelength. Through this method, we can get the transmission intensity ratio between modes at each port along wavelength. We calculated the wavelength of incident lights from the intensity ratio at each port using an optimization method. Our 12-channel mode decomposition spectrometer has a resolution of wavelength about 0.16 nm and can distinguish the two and more mixed wavelengths of the incident at the same time. 우리는 광 회로를 위한 on-chip nanostructure를 design하고 simulation 하였다. 우리는 silicon-on-insulator를 이용하여 단일 간섭계, 광대역 간섭계 그리고 single-port 간섭계에서 coherent perfect absorption을 제안하고 입증하였다. 또한, Parity-Time 구조를 이용하여 microring 간섭계에서 reverse circulation mode를 제안하고 입증하였다. 그리고 추가로, 우리는 진피 속 혈관 pattern의 imaging과 태양광 발전을 위한 linear Fresnel lens를 design하고 실험적으로 성능을 측정하였다. 마지막으로 inverse design을 이용하여 silicon-on-insulator에서 작동하는 12-channel mode decomposition spectrometer를 design 하였다. 시간 역행 lasing, coherent perfect absorber는 특정 파장의 빛을 완벽하게 흡수하기 때문에, sensor, 광 switch, 변조기 그리고 logic gate와 같은 광 회로를 위한 응용에 큰 잠재력이 있다. 일반적으로 coherent perfect absorber는 높은 Q-값을 가진 단일 간섭계를 사용한다. 그 결과, 흡수 spectrum은 매우 좁은 선폭을 가진다. 따라서 단일 간섭계에서의 CPA는 온도와 같은 환경적 변화에 민감하다. 매우 좁은 bandwidth는 sensor에 활용될 수 있지만, on-chip 광 회로에서는 까다로울 수 있다. 다층 구조를 사용하면 coherent perfect absorption에 기초한 넓은 완전 흡수 파장 대역을 얻을 수 있다. 1550 nm에서 silicon과 silicon dioxide는 빛을 흡수하지 않기 때문에, CPA는 광학적 흡수로 발생할 수 없다. 이는 scatterer를 사용하여 광학적 흡수 대신 광학적 손실로 전환하여 해결할 수 있다. 이는 coherent perfect loss라 부른다. Even 대칭을 가진 coherent perfect loss mode는 perfect magnetic conductor 경계를 사용하여 single-port에서 작동할 수 있다. Perfect magnetic conductor 역할을 하는 경계는 광결정에서 반사의 위상 전이를 조절하여 실현될 수 있다. 우리는 on-chip에서 광결정 nanobeam을 이용하여 단일 간섭계, 광대역 간섭계 그리고 single-port 간섭계에서 coherent perfect loss를 입증하였다. 우리는 3차원 finite element method를 사용하여 simulation을 진행하였다. 단일 간섭계를 사용하여, 우리는 1550 nm의 파장에서 완전한 손실, coherent perfect loss를 성취하였다. 이때, 손실의 full width at half max (FWHM)는 1.76 nm가 된다. 다양한 air-hole 크기로 구성된 다층 구조를 사용하여, 광대역 간섭계에서 coherent perfect loss 또한 성취되었다. Air-hole 크기가 달라지면 구조의 유효 굴절률이 달라진다. 따라서 광결정을 이용하여 다층 구조를 on-chip 상에 구현할 수 있다. 완전 손실 영역은 1549 nm과 1551 nm 사이로 넓어졌다. 우리는 이 완전 흡수 영역이 단일 간섭계에서의 coherent perfect loss보다 더 넓어졌다는 것을 확인하였다. 이때의 FWHM는 8.97 nm가 된다. 또한, 우리는 간단한 광결정 nanobeam의 반사 위상에 대한 성질을 이용하여 광결정 경계를 설정하고, 이를 최적화하여 perfect magnetic conductor-like 경계를 성취하였다. 광결정을 이용하여, 한쪽 면은 perfect magnetic conductor-like 경계를 다른 한쪽은 coupler가 위치한 Fabry-Perot 간섭계를 design 하였다. Coherent perfect loss의 손실 조건을 맞추기 위하여, coupler의 lattice constant를 조절하였다. 최적화된 coupling 손실에서, 우리는 single-port 구조에서 반사가 0이 되는 점을 찾았다. 이는 coherent perfect loss에 해당한다. 우리는 numerical analysis와 finite element method를 사용하여 Parity-Time 구조를 가진 microring 간섭계에서 reversed circulation mode를 제안하고 입증하였다. Reversed circulation mode에서, 일반적인 microring 간섭계와 다르게 빛이 microring 구조 내에서 입사 방향과 반대 방향으로 순환한다. 이는 free spectral range를 넘어선 넓은 파장 영역에서 유일하다. 우리는 directional coupling analysis와 transfer matrix를 사용하여, reverse circulation mode의 원리를 분석하였다. Finite element method를 사용하여 4-port Parity-Time microring 간섭계에서의 reverse circulation mode를 입증하였다. 우리가 제안한 reversed circulation mode는 optical logic gate, amplifying mirror 그리고 filter와 같은 optical circuit을 위한 응용의 잠재력이 있다. 피부 내부의 정맥 pattern은 개인마다 고유한 pattern을 가지고 있어 생체 인식에 사용될 수 있다. 하지만 피부에 의한 빛의 산란 때문에, lens와 같은 장비 없이 imaging 하는 것은 어렵다. 이는 커다란 detection system이 필요하다. 우리는 image scanning을 사용하여 lens 없이 정맥 pattern을 imaging 하기 위한 새로운 방법을 제시하였다. 우리는 simulation 상에서 image scanning 방법이 명확한 image를 제공한다는 것을 입증하였다. 추가로 더 작은 광원과 감지기는 더 명확한 image를 제공한다는 것을 확인하였다. 2 μm의 scan pixel 크기에서, 우리는 cross correlation G 값이 0.74에 도달한다는 것을 확인하였다. 우리는 태양광 발전을 위한 선형 Fresnel lens를 design 하였다. 최적화된 선형 Fresnel lens는 100 (W)×100 (H)×5 (T) mm^3의 크기를 가진다. 초점 거리는 100 mm이며 500 μm의 고정된 pitch 크기를 가진다. 최대 초점 효율은 대략 91%에 도달하였고, 초점 크기의 half width at half maximum (HWHM) 값은 130 μm를 가진다. 우리는 또한 입사광의 파장, 입사각, draft 각 그리고 감지기의 위치에 따른 선형 Fresnel lens의 성능을 조사하였다. Simulation 결과 입사각 1°, draft 각 10° 그리고 감지기의 위치가 최대 ±10 mm일 때, 85% 이상의 초점 효율을 얻는다는 것을 확인하였다. 우리는 injection molding process를 사용하여 선형 Fresnel lens를 제작하였고, 제작된 lens가 최대 88%의 초점 효율과 2.5 mm의 초점 크기의 FWHM 값을 가진다는 것을 확인하였다. 마지막으로 우리는 silicon-on-insulator에서 inverse design을 이용하여 입사 mode에 따라 경로를 분해할 수 있는 mode 분광기를 구현하였다. Inverse design은 전기장 분포에 따라 구조를 최적화할 수 있는 design 방법이다. 이를 통하여 소형 device에 파장 분광기, mode converter 그리고 mode 분광기를 제작할 수 있다. 우리는 다중 mode를 전파할 수 있는 7.5 μm 너비의 input waveguide와 단일 mode만을 12개의 전파하는 500 nm 너비의 output waveguide 그리고 10×10 μm^2의 design 영역으로 구성된 mode 분광기를 구성하였다. 우리는 최적화 유효 굴절률을 이용하여 2차원상에서 inverse design을 수행하였고, 3차원 finite-difference time-domain simulation을 통하여 구조의 작동을 확인하였다. 우리의 구조는 최대 39%까지의 투과효율을 보였으며 11% 이하의 mode 혼선을 보여준다. Scattering 영역을 추가함으로, 빛은 파장에 따라 다른 비율의 다중 mode로 혼합이 될 수 있다. 이를 통하여 파장에 따라 각 port로 투과되는 mode들 사이의 다른 intensity 비율을 구할 수 있다. 이 결과값을 최적화 계산을 통하여 입사광의 파장을 추정할 수 있다. 즉, 각 port에서 얻은 intensity 비율을 통하여 입사되는 빛의 파장을 구할 수 있다. 우리는 mode 분광기에 scattering 영역을 추가하여, 12-channel mode decomposition spectrometer를 구현하였다. 우리의 구조는 trial 파장에 대하여, 정확한 값의 파장을 식별할 수 있고 또한 파장의 입사광이 둘 이상 섞여 있을 때, 최대 약 0.16 nm의 해상도를 가지고 파장을 구분해 낼 수 있다.

Age at first childbirth, number of childbirths, and hypertriglyceridemia in later life

최혜린 Graduate School, Yonsei University 2020 국내박사

KOSPI200 주가지수 옵션에서의 Put-Call Parity를 이용한 차익거래 전략 및 시장의 안정성에 대한 연구

경강수 가톨릭대 대학원 2005 국내석사

Put-call parity is a relationship, first identified by Stoll, that must exist between the prices of European put and call options that both have the same underling, strike price and expiration date. The relationship is derived using arbitrage arguments. If put-call parity dose not hold, there are arbitrage opportunities. The test of put-call parity used date on option prices taken from trades in KOSPI200 sock index option. The conclusion is that put-call parity does not hold and arbitrage opportunities were available. However, when transactions costs were taken into account, there were no profitable opportunities frequently.

Influence of Dietary Nutrition Content on Milking Frequency, Animal Production and Blood Metabolites of Lactating Cows in Automatic Milking System

이동석 경상대학교 대학원 2016 국내박사

EXPERIMENT I. Feeding starchy pellets with high inert fat increased milk performance of lactating cows in free-stall house with automatic milking system. The provision of concentrates to cows visiting automatic milking system should be strictly controlled because, the over-suppling of starch rich diets within a short time could reduce fiber digestion by ruminal bacteria. Altering the composition of concentrate diets by replacing a starch-rich grain with a by-product higher in digestible NDF is one of strategies to increase the amount of concentrate diets supplied in the AMS without negative effect on rumen fermentation and milking attendance. However, feeding excessive amounts of digestible NDF decreases feed intake and lowers the feed efficiency due to reduced microbial protein synthesis. Therefore, hypothesis to be tested in this study that starchy concentrate diets with high inert fat could replace concentrate diets high in digestible neural detergent fiber to improve animal performance without adverse effect on carbohydrate and nitrogen metabolism in the rumen and the voluntarily cows’ motivation to visit a milking unit. Thirty-eight cows were blocked according to age, milk yield, and days in milk into two groups then divided randomly into 2 dietary treatment groups. The 2 type of pellets differed in the energy density and composition of the concentrate diets fed in the AMS : for the control gorups, the pellets rich in digestible NDF were composed of 30.7% starchy grains and 15.0% wheat bran, Whereas for the treatment groups, starchy pellets with high inert fat contained 38.0% starchy grains with inert fat (2%) as wheat bran replacements. Both groups resulted in similar numbers of voluntary milkings (4.00 vs 4.01 times per cow per day, respectively). Accordingly milking interval was similar in both treatments (6.59 vs. 6.74 h, respectively). When control diets (pellets rich in digestible NDF) was replaced by treatment diets (starchy pellets with high inert fat), although there were not significantly, total fat intake and NSC intake tended to be increased (P = 0.245 and P = 0.172). Although there were not significantly, Milk yield (kg of milk/cow per d) of treatment groups tended to be higher than those of the control group (31.44 and 28.53 kg/d, respectively, P = 0.186). The average concentrations of milk fat and milk protein were similar in the both groups. However, the production of milk fat, 4% FCM, and ECM were higher (P = 0.070, P = 0.101, and P = 0.125) trend in the treatment group than in the control group. The results suggest that starch pellets with high inert fat supplied through the automatic concentrate-feeder might improve milk yield, FCM, ECM, and milk fat yield without negative effect on rumen fermentation and the voluntarily cows’ motivation to visit a milking unit. (Key words : AMS, NDF, inert fat, FCM, ECM, NSC) EXPERIMENT II. Feeding starchy pellets with high inert fat increased milk performance of lactating cows in loose house with automatic milking system. The provision of concentrates to cows visiting automatic milking system should be strictly controlled because, the over-suppling of starch rich diets within a short time could reduce fiber digestion by ruminal bacteria. Altering the composition of concentrate diets by replacing a starch-rich grain with a by-product higher in digestible NDF is one of strategies to increase the amount of concentrate diets supplied in the AMS without negative effect on rumen fermentation and milking attendance. However, feeding excessive amounts of digestible NDF decreases feed intake and lowers the feed efficiency due to reduced microbial protein synthesis. Therefore, hypothesis to be tested in this study that starchy concentrate diets with high inert fat could replace concentrate diets high in digestible neural detergent fiber to improve animal performance without adverse effect on carbohydrate and nitrogen metabolism in the rumen and the voluntarily cows’ motivation to visit a milking unit. Thirty cows were blocked according to age, milk yield, and days in milk into two groups then divided randomly into 2 dietary treatment groups. The 2 type of pellets differed in the energy density and composition of the concentrate diets fed in the AMS : for the control gorups, the pellets rich in digestible NDF were composed of 27.0% starchy grains and 15.0% wheat bran, Whereas for the treatment groups, starchy pellets with high inert fat contained 41.0% starchy grains with inert fat (3%) as wheat bran replacements. Milking frequency were higher (P < 0.05) when cows received the treatment diets (2.93 milking/day) than when cows received the control diets (2.38 milking/day). As a result, the milking interval of the treatment group tended to faster than this of the control group (8.76 and 10.77 h, respectively, P ≤ 0.05). Although there were no significant differences both the treatments, the NEl balance trend to be negative in treatment groups (NEl balance: -0.71 Mcal/d) compare with control groups (NEl balance: 1.99 Mcal/d). The result showed that the motive for visiting the milking unit might be increased to compensate for the lack of nutrients, when nutrients that can be obtained from the basal mixture were limited. Although there were not significantly, ME and NEl intake tended to be higher in cows received the treatment diets (ME intake: 57.56 Mcal/d, and NEl intake: 36.36 Mcal/d) than in cows received the control diets (ME intake: 52.10 Mcal/d, and NEl intake: 33.02 Mcal/d). Fat intake, NEm and NEg were significantly higher in treatment group compare with control groups. This result suggest that increasing the energy density of the concentrate diets fed in the milking unit with inert fat is one of strategies to increase total energy intake without adverse effect of rumen fermentation and animal performance. Although there were not significantly, the average milk yields of cows in the treatment groups tended to be higher than those of the cows in control group (34.44 and 28.28 kg/d, respectively, P = 0.066). The average concentrations of milk fat and milk protein were similar in the 2 groups; however, the production of milk fat, 4% FCM, and ECM were significantly higher (P ≤ 0.05) in the treatment group than in the control group. The results suggest that starch pellets with high inert fat supplied through the automatic concentrate-feeder could improve milk yield, FCM, ECM, and milk fat yield without negative effect on rumen fermentation and the voluntarily cows’ motivation to visit a milking unit. (Key words : AMS, NDF, inert fat, FCM, ECM, NSC, Milking frequency) EXPERIMENT III. Influence of dietary protein content on milking frequency, animal production and blood metabolites of mid-lactating cows in an automatic milking system Twenty-seven Holstein cows were blocked by age, milk yield, and days in milk into three groups [concenterate diets containing a low (LPD: CP 199g/kg of DM), medium (MPD: CP 210g/kg of DM) or high (HPD: CP 221g/kg of DM) protein according to crude protein levels of concentrate feed fed through AMS] then randomly assigned to one of three diet sequences. The experiment lasted 30 days composed of 15 days for diet adaptation and 15 days for data collection. The concentrate diets were distinguished by composition of corn grain, vegetable oil, and soybean meal. There was no significant in the numbers of voluntary milkings (3.02, 3.25, and 2.89 per cow per day, respectively). Increasing crude protein levels of concentrate diets from MPD to HPD significantly increased (P < 0.05) total dietary protein intake, RUP supply, RUP balance, or MP balance, while milk protein concentration was significantly decreased (P < 0.05). There was a major decline in the efficiency of nitrogen utilization for milk production in animals (HPD) fed 4.45 kg of CP per day throughout mid-lactation, compared with those (LPD and MPD) fed 4.16 and 4.13 kg of CP per day. Additionally animals (MPD: MP balance of -68.4 g/d) receiving 4.13 kg of CP per day showed no detrimental effects on nitrogen efficiency, compared with animals (HPD: MP balance of 188.7 g/d) fed 4.45 kg of CP per day. This result showed that nitrogen efficiency could be improved with cows fed the MP-deficient diets. Despite the high MP balance and RUP balance in HPD compare with MPD, there was no further increase milk yield and milk protein content. Additionally, increasing crude protein levels of concentrate diets from MPD to HPD significantly decreased (P < 0.05) total essential amino acids in MP. This result may be attributed to the relatively more unbalanced amino acid composition in HPD compare with MPD, especially Lys and Met in MP, which reduced the efficiency of MP utilization for milk protein synthesis and nitrogen efficiency. Furthermore, increasing total protein intake resulted in comparable increases in BUN and decreases in N efficiency. Improving efficiency and economic sustainability of dairy cow production systems is critical, as is reducing the environmental impact of intensive dairy farming. These results suggest that reducing dietary protein concentration in mid lactation improves the efficiency of use of dietary nitrogen with no detrimental effects on animal performance. (Key words : AMS, CP, RDP, RUP, MP, Total EAA, Milking frequency, dLys, dMet, nitrogen effiency, BUN) EXPERIMENT IV. Influence of dietary protein content on milking frequency, animal production and blood metabolites of late-lactating cows in an automatic milking system Twenty-seven Holstein cows were blocked by age, milk yield, and days in milk into three groups [concenterate diets containing a low (LPD: CP 199g/kg of DM), medium (MPD: CP 210g/kg of DM) or high (HPD: CP 221g/kg of DM) protein according to crude protein levels of concentrate feed fed through AMS] then randomly assigned to one of three diet sequences. The experiment lasted 30 days composed of 15 days for diet adaptation and 15 days for data collection. The concentrate diets were distinguished by composition of corn grain, vegetable oil, and soybean meal. The milking frequency were significantly affected by dietary treatment, with the lowest value for HPD (Milk frequency of 2.69 milking/day), followed by LPD (Milk frequency of 2.89 milking/day), and the highest for MPD (Milk frequency of 3.50 milking/day). Although there were no significant differences among the treatments, the MP balance trend to be negative in MPD groups (MP balance: -32.49 g/d) compare with LPD groups (MP balance: 95.72 g/d) and HPD (MP balance: 97.83 g/d). The result showed that the motive for visiting the milking unit might be increased to compensate for the lack of nutrients, when nutrients that can be obtained from the basal mixture were limited. The digestible Lys and Met in MP fed to cows through the feeding line and automatic concentrate-feeder were calculated to be 6.09 and 1.85, 6.07and 1.84, and 6.09% and 1.84% of MP per day for LPD, MPD, and HPD respectively, both dLys and dMet lower than recommended by NRC (2001) in all treatment. There were not significant total essential amino acids (Total EAA) in MP. Due to the increase of milk yield, total milk fat yields (1.35 kg/day versus 1.51 kg/day, P < 0.05) increased with increasing total protein intake (LPD; 4.00 kg/day versus MPD; 4.20 kg/day of CP respectively). There were no increase milk yield, milk protein content and the efficiency of nitrogen utilization for milk production by increasing crude protein levels of concentrate diets supplied through AMS during late-lactation. These results may be attributed to the unbalanced amino acid composition in all treatment, especially Lys and Met in MP, which reduced the efficiency of MP utilization for milk protein synthesis and nitrogen efficiency. (Key words : AMS, CP, RDP, RUP, MP, Total EAA, Milking frequency, dLys, dMet, nitrogen effiency, BUN) EXPERIMENT V. Modeling lactation curve of dairy cows using biometric data collected from farms equipped with automatic milking systems in South Korea The milking characteristics during the course of whole-lactation in AMS were studied using biometric data collected from farms equipped with automatic milking. Additionally, descriptions of entire lactations were investigated using two mathematical lactation models, comprising the performance with the Wood equation (the most popular empirical model of the lactation curve) and the Dijkstra equation (the representative mechanistic model) by fitting these equations to daily milk records collected from farms. Ten farms using a free-flow traffic (Lely, Maassluis, the Netherlands) automatic milking system in South Korea were included in the study. Three farms had 2 AMS per farm, and 7 farms had 1 AMS per farm. However, each farm only permitted a cow’s access to 1 AMS each. The data used in this study 27,145, 16,470, and 17,080 daily records taken from 89 first, 55 second, and 56 third and greater parity cows, respectively. The shape of milk frequency was similar with lactation curves regardless of parity during the course of whole-lactation. Although the changes were small, milk frequency were significant increased (P <0.001) with increase in parity order during early lactation. Lactation data showed a significantly lower initial yield (P < 0.001) for the first parity than for second and greater parity cows, as expected. Also, a significantly lower peak yield (P < 0.001), a later time to peak (P < 0.05) and a significantly greater persistency (P < 0.001) were obtained for first parity cows than for those in later parities. Lactation persistency was significantly poorer (P < 0.001) in second and greater parity cows then in first parity cows (Slopes of yield decline of -0.085, -0.070, and -0.027, respectively). The residual distribution from Wood and Dijkstra fluctuated throughout the lactation between positive and negative. The distribution pattern from these functions underestimated after calving between DIM 1 to 10, but overestimated around peak yield in greater parity groups. Based on criteria to measure goodness of fit, statistical evaluation revealed that MSE and RMSE had significant differences from both functions in all parity groups. Therefore given the results, the Dijkstra function performed more accurately in fitting individual lactation than Wood. (Key words : AMS, Lactation curve, Dijkstra, Wood, Parity, Peak yield, Lactation persistency, Milking frequency) 실험 I. 계류식 우사(free-stall barn)의 자동착유시스템을 이용한 보호 지방 첨가가 젖소의 착유패턴 및 혈액 성상에 미치는 영향 자동착유시스템에서 젖소가 짧은 시간 동안 전분 위주의 배합사료를 과량 섭취할 경우 반추위 미생물에 의한 소화율이 떨어지게 된다 (Miron et al., 2004a, b). 때문에 착유 시설의 배합사료의 공급은 엄격히 제어되어야 한다. 곡류 위주의 전분질 배합사료를 가소화 NDF 함량이 높은 부산물로 대체하여 사료의 구성 성분을 변경하는 것은 반추위 발효와 젖소의 착유 행동 패턴에 부정적인 영향 없이 착유 시설에서 공급하는 배합사료의 양을 증량시키는 전략 중 하나이다 (Halachmi et al., 2008). 하지만 과도한 NDF의 공급은 사료 섭취량 감소와 반추위 미생물 합성의 감소로 인해 사료의 이용 효율이 낮아 질 수 있다. 따라서 본 실험의 가설은 밀기울처럼 가소화 NDF 함량이 높은 부산물을 주요 원료로 하는 배합사료를 높은 수준의 보호 지방을 첨가한 전분질 배합사료로 대체 할 경우 자발적으로 착유시설을 방문하려는 젖소의 의지와 반추위 단백질 및 탄수화물대사에 부정적인 영향 없이 생산성을 향상 시킬 수 있다는 것이다. 실험을 위해 착유 젖소 38두 (초산 18두, 경산 20두)를 공시하였다. 자동 착유 시스템에서 착유시설 안에서 공급되는 배합사료의 종류에 따라 control 그룹과 treatment 그룹으로 나누었고, 각 그룹 당 15두씩 고르게 배치하였다. 실험에 사용된 배합사료는 착유시설에서 제공하는 배합사료의 에너지 구성과 에너지 농도에 따라 30.70%의 곡류, 15.0%의 밀기울이 포함된 control 배합사료 (가소화 NDF가 높은 배합사료)와 보호 지방과 곡류의 첨가 비율을 조절하여 control 배합사료의 밀기울을 일부 대체 하여 제조한 treatment 배합사료 (보호지방 첨가 전분질 배합사료)를 사용하였다. 두 실험 배합사료의 조단백질은 20.0%/DM으로 고정하였다. 착유 횟수는 treatment 그룹 (4.00 milking/day)과 control 그룹 (4.01 milking/day)이 유사하다는 것을 알 수 있다. 이로 인해 착유 간격 또한 처리구간에 차이가 없었다 (6.59 와 6.74 h). 비록 유의적인 차이는 없었지만 총 지방 섭취량과 NSC 섭취량은 treatment 그룹이 control 그룹 보다 높은 경향을 보였다 (P = 0.245, 그리고 P = 0.172). 비록 유의적인 차이는 없었지만 일일 유량은 treatment 그룹이 control 그룹 보다 높은 경향을 보였다 (31.44 그리고, 28.53 kg/d, P = 0.186). 또한 1회 착유 유량도 treatment 그룹이 control 그룹 보다 높은 경향을 보였다 (7.73, 그리고 6.81 kg/d, P = 0.204). 유지율과 유단백질 함량은 두 그룹이 유사하였다. 하지만 유지방 생산량, 4% FCM 및 ECM은 treatment 그룹이 control 그룹보다 높은 경향을 보였다 (P = 0.070, P = 0.101 그리고, P = 0.125). 이러한 실험 결과는 높은 수준의 보호 지방을 첨가한 전분질 배합사료는 반추위 발효 성상과 착유 패턴에 부정적인 영향 없이 FCM, ECM, 및 유지방 생산량을 증가시킬 수 있다는 것을 보여준다. 실험 II. 방사식 우사(loose barn)의 자동착유시스템을 이용한 보호지방 첨가 전분질 펠렛사료의 급여가 젖소의 생산성에 미치는 영향 자동착유시스템에서 젖소가 짧은 시간 동안 전분 위주의 배합사료를 과량 섭취할 경우 반추위 미생물에 의한 소화율이 떨어지게 된다 (Miron et al., 2004a, b). 때문에 착유 시설의 배합사료의 공급은 엄격히 제어되어야 한다. 곡류 위주의 전분질 배합사료를 가소화 NDF 함량이 높은 부산물로 대체하여 사료의 구성 성분을 변경하는 것은 반추위 발효와 젖소의 착유 행동 패턴에 부정적인 영향 없이 착유 시설에서 공급하는 배합사료의 양을 증량시키는 전략 중 하나이다 (Halachmi et al., 2008). 하지만 과도한 NDF의 공급은 사료 섭취량 감소와 반추위 미생물 합성의 감소로 인해 사료의 이용 효율이 낮아 질 수 있다. 따라서 본 실험의 가설은 밀기울처럼 가소화 NDF 함량이 높은 부산물을 주요 원료로 하는 배합사료를 높은 수준의 보호 지방을 첨가한 전분질 배합사료로 대체 할 경우 자발적으로 착유시설을 방문하려는 젖소의 의지와 반추위 단백질 및 탄수화물대사에 부정적인 영향 없이 생산성을 향상 시킬 수 있다는 것이다. 실험을 위해 착유 젖소 30두 (초산 10두, 경산 20두)를 공시하였다. 자동 착유 시스템에서 착유시설 안에서 공급되는 배합사료의 종류에 따라 control 그룹과 treatment 그룹으로 나누었고, 각 그룹 당 15두씩 고르게 배치하였다. 실험에 사용된 배합사료는 착유시설에서 제공하는 배합사료의 에너지 구성과 에너지 농도에 따라 27.0%의 곡류, 15.0%의 밀기울이 포함된 control 배합사료 (가소화 NDF가 높은 배합사료)와 보호 지방과 곡류의 첨가 비율을 조절하여 control 배합사료의 밀기울을 일부 대체 하여 제조한 treatment 배합사료 (보호지방 첨가 전분질 배합사료)를 사용하였다. 두 실험 배합사료의 조단백질은 20.0%/DM으로 고정하였다. 착유 횟수는 treatment 그룹 (2.93 milking/day)이 control 그룹 (2.38 milking/day) 보다 유의적으로 높았다 (P < 0.05). 이로 인해 착유 간격은 treatment 그룹이 control 그룹 보다 유의적으로 빨랐다 (8.76 그리고, 10.77 h, P < 0.05). 비록 유의적인 차이는 없었지만 NEl 균형은 treatment 그룹 (NEl balance: -0.71 Mcal/d)이 control 그룹 (NEl balance: 1.99 Mcal/d) 보다 낮은 경향을 보이며, 음의 에너지 균형이었다. 이러한 실험결과는 Feed fence에 제공되는 기초 사료로부터 얻을 수 있는 일부 영양소가 제한되어 부족한 영양소를 보충하기 위해 자발적으로 착유 시설을 방문하려고 하는 젖소의 의지가 더욱 강해진 것일지도 모른다. ME 섭취량, 그리고 NEl 섭취량은 treatment 그룹 (ME intake: 57.56 Mcal/d 그리고, NEl intake: 36.36 Mcal/d)이 control 그룹 (ME intake: 52.10 Mcal/d 그리고, NEl intake: 33.02 Mcal/d) 보다 높은 경향을 보였다. 또한 지방 섭취량, NEm 섭취량과 NEg 섭취량은 control 배합사료에서 treatment 배합사료로 대체 될 때 유의적으로 증가하였다. 이러한 실험결과는 보호지방을 이용하여 착유시설에서 제공하는 배합사료의 에너지 수준을 증가시키는 방법은 반추위 발효와 착유 패턴에 부정적인 영향 없이 에너지 섭취량을 증가시킬 수 있다는 것을 알 수 있다. 비록 유의적인 차이는 없었지만 평균 유량은 treatment 그룹이 control 그룹 보다 높은 경향을 보였다 (34.44 그리고, 28.28 kg/d, P = 0.066). 유지방과 유단백질 함량은 두 처리구 모두 유사하였다. 하지만 유지방 생산량, 4% FCM, 그리고 ECM은 treatment 그룹이 control 그룹 보다 유의적으로 높았다 (P < 0.05). 이러한 실험 결과는 높은 수준의 보호 지방을 첨가한 전분질 배합사료는 반추위 발효 성상과 착유 패턴에 부정적인 영향 없이 FCM, ECM, 및 유지방 생산량을 증가시킬 수 있다는 것을 보여준다. 실험 III. 자동착유시스템에서 자동공급 장치를 통해 제공되는 배합사료의 단백질 수준이 비유 중기 젖소의 착유 횟수, 생산성 및 혈액대사에 미치는 영향 실험을 위해 착유 젖소 27 마리를 공시하였고, 자동착유시스템에서 자동급여기를 통해 제공되는 배합사료의 단백질 수준에 따라 3개의 실험구 [배합사료의 CP 수준(LPD: CP 199g/kg of DM, MPD: CP 210g/kg of DM 그리고 HPD: CP 221g/kg of DM)]로 나누었으며 각 실험구 마다 산차, 착유일수, 및 유량을 고려하여 7두씩 배치하였다. 실험 배합사료는 옥수수와 대두박의 혼합비를 조절하여 고단백질, 중단백질 그리고 저단백질의 배합사료를 제조하였고, 산차와 비유단계에 따라 단백질 수준이 다른 실험 펠렛 사료를 설정된 프로그램에 따라 급여하였다. 실험 기간은 30일로 사료의 적응 기간 15과 자료 수집기간을 15일로 두었다. 착유 횟수는 처리구간에 유의적 차이는 없었다. 단백질 수준이 중단백질 (MPD)에서 고단백질 (HPD)로 증가할 때 총단백질 섭취량, RUP 공급량, RUP 균형, 및 MP 균형은 유의적으로 증가하였고( (P < 0.05), 반면에 유단백질 함량은 유의적으로 감소하였다 (P < 0.05). 실험 기간 동안 HPD 그룹의 질소 이용 효율은 MPD와 LPD 그룹에 비해 유의적으로 낮았다. 또한 총 단백질 섭취량 4.13 kg/CP/day의 중단백질 급여구 (MPD: MP balance of -68.4 g/d)는 총 단백질 섭취량 4.45 kg/CP/day의 고단백질 급여구 (HPD: MP balace of 188.7 g/d)에 비해 단백질 이용 효율에 부정적인 영향이 없었다. 이러한 실험 결과는 MP가 결핍된 사료를 섭취한 착유 젖소에서 질소의 이용 효율이 더 높다는 것을 알 수 있다. 중단백질 급여구 (MPD: MP balace of -68.4 g/d 와 RUP balace of -82.14 g/d) 에 비해 고단백질 급여구 (HPD)는 더 높은 MP 균형 (MP balance of 188.7 g/d) 과 RUP 균형 (RUP balance of 227.05 g/d)에도 불구하고 유 단백질과 유량의 증가는 없었다. 중단백질 (MPD)에서 고단백질 (HPD)로 단백질 수준이 증가할 때 MP 내 필수 아미노산 함량과 digestble 아미노산 (dArg, dHis, dLeu, dLeu, dMet, 그리고 dThr)들의 함량은 유의적으로 감소하였다. 이러한 결과는 중단백질 급여구 (MPD)에서 보다 고단백질 급여구 (HPD)에서 불균형한 아미노산 조성 (특히 Met와 Lys)으로 인해 질소의 이용 효율과 단백질 합성을 위한 MP의 이용 효율이 감소한 결과 일지도 모른다. 또한 단백질 섭취량의 증가는 BUN 함량을 증가시키고, 질소의 이용 효율 감소를 야기하였다. 이러한 실험 결과를 종합해 볼 때 단백질 함량과 단백질 섭취량의 감소는 비유 중기 착유 젖소의 생산성에 부정적인 영향 없이 질소의 이용효율을 향상시킬 수 있다는 것을 보여준다. 실험 IV. 자동착유시스템에서 자동공급 장치를 통해 제공되는 배합사료의 단백질 수준이 비유 말기 젖소의 착유 횟수, 생산성 및 혈액대사에 미치는 영향 실험을 위해 착유 젖소 27 마리를 공시하였고, 자동착유시스템에서 자동급여기를 통해 제공되는 배합사료의 단백질 수준에 따라 3개의 실험구 [배합사료의 CP 수준(LPD: CP 199g/kg of DM, MPD: CP 210g/kg of DM 그리고 HPD: CP 221g/kg of DM)]로 나누었으며 각 실험구 마다 산차, 착유일수, 및 유량을 고려하여 7두씩 배치하였다. 실험 배합사료는 옥수수와 대두박의 혼합비를 조절하여 고단백질, 중단백질 그리고 저단백질의 배합사료를 제조하였고, 산차와 비유단계에 따라 단백질 수준이 다른 실험 펠렛 사료를 설정된 프로그램에 따라 급여하였다. 실험 기간은 30일로 사료의 적응 기간 15과 자료 수집기간을 15일로 두었다. 착유 횟수는 처리구들 간에 유의적인 차이가 있었고, 고단백질 급여구 (HPD: 2.69 milking/day)로 가장 낮았으며, 그 다음이 저단백질 (LPD: 2.89 milking/day), 그리고 중단백질 (MPD: 3.50 milking/day) 순이었다. 비록 처리구간에 유의차는 없었지만 중단백질 급여구 (MP balance of -32.49 g/d)의 MP 균형은 고단백질 급여구 (MP balance of 97.89 g/d) 와 저단백질 급여구 (MP balance of 95.72 g/d)에 비해 낮은 경향을 보였고, 음의 MP 균형이었다. 이러한 실험결과는 Feed fence에 제공되는 기초 사료로부터 얻을 수 있는 일부 영양소가 제한되어 부족한 영양소를 보충하기 위해 자발적으로 착유 시설을 방문하려고 하는 젖소의 의지가 더욱 강해진 것일지도 모른다. MP 내 digestible Lys 과 Met 함량은 저단백질 급여구가 6.09%와 1.85% 이였고, 중단백질 급여구가 6.07%와 1.84% 였으며, 고단백질 급여구가 6.09% 와 1.84% 이였다. Digestible Lys 과 Met 함량 모두 모든 처리구에서 NRC가 제시한 요구량 보다 낮았다. MP 내 필수 아미노산 총량은 처리구간에 유의차가 없었다. 저단백질 (4.0 kg/day) 에서 중단백질 (4.20 kg/day)로 총 단백질 섭취량이 증가할수록 유량의 증가로 인해 유지방 생산량은 유의적으로 증가하였다 (1.35 kg/day, 그리고 1.51 kg/day, P < 0.05). 비유 말기 유량, 유단백질 함량, 및 질소 이용효율은 처리구간에 유의적인 차이가 없었다. 이러한 결과는 모든 처리구에서 불균형한 아미노산 조성 (특히 Met와 Lys)으로 인해 질소의 이용 효율과 단백질 합성을 위한 MP의 이용 효율이 감소한 결과 일지도 모른다. 실험 V. 자동 착유시스템 도입 목장으로부터 수집한 젖소의 생체정보를 활용한 착유 젖소의 비유 곡선 모델링에 관한 연구. 자동 착유시스템에서 수집한 젖소의 생체 정보를 활용하여 젖소의 산차별 비유기 전 기간 동안의 비유 특성 관해 조사하고, 대표적인 비유 곡선 모델인 Wood 모델 (the most popular empirical model of the lactation curve) 과 Dijkstra 모델 (the representative mechanistic model)의 비유곡선 추정에 대한 각 모델의 적합도를 평가하였다. 자동착유시스템 (Lely, Maassluis, The Netherlands)이 설치된 10 곳의 착유 농장으로 부터 총 착유우 738두의 착유데이터를 수집하였다. 이 중 젖소 모델 연구를 위해 자동착유시스템에서 비유 1일부터 305일까지 착유가 완료된 착유우 200두 (1산 89, 2산 55, 그리고 3산 이상 56두)로 부터 60,695 건 (1산 27,145건, 2산 16,470건, 3산 17, 080건)의 일일 착유 데이터를 이용하였다. 자료는 자동 착유시스템을 도입하여 운영한 기간이 평균 13개월 이상인 농가로부터 수집하였고, 최대 22개월 이상, 최소 12 개월 이상 자동 착유 시스템을 운영한 농가로부터 수집하였다. 젖소의 착유 횟수는 비유 곡선과 유사한 경향을 보이며 비유 일수가 경과 할수록 감소 서서히 감소하였다. 비록 비유 초기 동안의 평균 착유 횟수의 차이는 크지 않았지만, 산차가 증가할수록 유의적으로 증가하였다 (P < 0.001). 비유 개시일의 평균 유량과 평균 최고유량은 경산우 그룹이 초산 그룹에 비해 유의적으로 높았다 (P < 0.001). 최고유량 도달 시기는 2산 이상 그룹이 초산 그룹에 비해 유의적으로 빨랐다 (P < 0.05). 하지만 비유 지속성은 경산우 그룹이 초산우 그룹보다 유의적으로 낮았다 (3산 이상: -0.085, 2산: -0.070, 그리고 1산: -0.027, P < 0.001). Wood 와 Dijkstra 모델의 임의 오차의 산포도는 비유 기간 동안 고르지 않은 경향을 보였다. 이 두 모델 모두 분만 후 1~10일 까지 실측치 보다 과소평가 되었다가, 최고 유량 도달 시기에 과대평가되는 경향을 보였다. 이상의 결과를 종합해 볼 때 초산우의 비유 곡선은 경산우에 비해 편평한 비유 곡선을 가지고, 최대유량이 낮으며, 비유지속성이 높은 특징이 있다. 반면 경산우는 초산우에 비해 최대유량이 높고, 비유 지속성이 낮다. 각 모델의 적합도 평가에서 MSE와 RMSE는 모든 산차에서 두 모델간의 유의적 차이가 있었고, Dijkstra 모델이 낮았다. 이러한 결과를 종합해 볼 때 Dijkstra 모델이 Wood 모델에 비해 비유곡선 추정의 적합도가 더 뛰어나다는 것을 알 수 있다.

Relationships among parity, sleep duration, and obesity according to the lifecycle of Korean women

안효정 서울시립대학교 도시보건대학원 2021 국내석사

Objectives: Physiological changes can be the result of various health problems in postpartum women. The aim of the present study was to address the relationships among parity, sleep duration, and obesity across the lifecycle of adult Korean women. Methods: Our study analyzed data from 10,081 participants who provided responses to the Korean National Health and Nutrition Examination Survey from 2010 to 2012. We divided the data into subgroups according to age [Group 1 (20-39 years), Group 2 (40-64 years), and Group 3 (≥65 years)], and parity (0, 1, 2, 3, and ≥4). We defined sleep duration as the average time spent sleeping per day. Obesity was defined as a body mass index (BMI) greater than 25.0kg/m². Abdominal obesity was defined as waist circumference (WC) greater than 85.0cm. The relationships among parity, sleep duration, and obesity were investigated using logistic and linear regression analyses. Results: Sleep duration in women younger than 65 years decreased with increased parity [Group 1 (β): -0.174±0.052, p: 0.001; Group 2 (β): -0.083±0.028, p: 0.003]. The risk of obesity increased in women aged 40 years of age and older [Group 2 (OR): 1.18 (1.09-1.29), p: <0.001; Group 3 (OR): 1.09 (1.02-1.16), p: 0.010]. The risk of abdominal obesity also increased in same age groups [Group 2 (OR): 1.22 (1.12-1.33), p: <0.001; Group 3 (OR): 1.11 (1.04-1.18), p: 0.001]. However, sleep duration and obesity demonstrated no significant relationship in any age group. Conclusions: Korean women are found to manifest a significant negative association between parity and sleep duration. Also, an increased risk of obesity is observed in women had more parity. 목적: 출산 후 여성이 겪는 생리학적 변화들은 다양한 건강 문제들이 발생하는 결과가 될 수 있다. 이 연구의 목표는 자녀 수, 수면시간, 그리고 비만 간의 관계를 한국 성인 여성의 생애주기에 따라 제시하는 것이다. 방법: 이 연구에서는 2010년부터 2012년까지 한국 국민건강영양조사에 참여한 10,081명을 연령에 따라 3개의 그룹(20-39, 40-64, 65세 이상)으로 세분화하여 분석하였다. 수면시간은 하루 평균 수면시간을 나타낸다. 비만은 체질량지수 ≥25kg/㎡로 정의하였고, 복부비만은 허리둘레 ≥85㎝로 정의하였다. 자녀 수와 수면시간, 비만 간의 관계는 로지스틱 및 선형회귀분석을 통해 조사하였다. 결과: 65세 미만 여성의 수면시간은 자녀 수가 많을수록 감소하였다 [그룹 1 (β): -0.174±0.052, p: 0.001; 그룹 2 (β): -0.083±0.028, p: 0.003]. 비만의 오즈(odds)는 40세 이상 여성에서 자녀 수가 많을수록 증가하였으며 [그룹 2 (OR): 1.18 (1.09-1.29), p: <0.001; 그룹 3 (OR): 1.09 (1.02-1.16), p: 0.010], 복부비만의 오즈 또한 동일한 연령대에서 증가하였다 [그룹 2 (OR): 1.22 (1.12-1.33), p: <0.001; 그룹 3 (OR): 1.11 (1.04-1.18), p: 0.001]. 하지만 수면시간과 비만의 유의미한 관계는 어느 연령대에서도 확인할 수 없었다. 결론: 한국 여성들에게 자녀 수와 수면시간 간의 부정적인 관계가 유의하게 나타났다. 또한, 자녀 수가 많은 여성은 비만의 위험이 증가하는 것을 관찰하였다.

한국의 구매력평가설 성립에 대한 연구

온심열 전북대학교 일반대학원 2020 국내석사

Purchasing Power Parity theory is one of the most influential theories of exchange rate determination, and often becomes an important basis for countries to formulate economic policies. The theory holds that the currency exchange rates of the two countries depend on the purchasing power of the two currencies. Since the theory was proposed, many countries have widely referred to this theory when formulating or reforming the exchange rate system. Whether or not the exchange rate policy of each country can be formulated based on the theory of purchasing power parity has become the focus of attention of all countries. However, economists have obvious differences on the correctness of purchasing power parity theory. There are a large number of theoretical and empirical literatures supporting and opposing the theory. This paper takes South Korea as the center, and the research target countries are developed countries (USA, United Kingdom, Canada) and developing countries (China) for empirical analysis. Choose the monthly exchange rate and CPI(Consumer Price Index) data from January 1980 to December 2018 between South Korea, the United States and the United Kingdom; South Korea and Canada use monthly exchange rate and CPI data from January 2000 to December 2018; South Korea and China use monthly exchange rate and CPI data from January 2006 to December 2018. Empirical analysis was performed using unit root and Johansen co-integration tests. The results of the empirical analysis are as follows: There is a co-integration relationship between South Korea/USA, South korea/United Kingdom, and South Korea/Canada. Purchasing power parity is established. However, only the Trace co-integration relationship exists in South Korea/China, but the Maximum Eigenvalue co-integration relationship does not exist. It cannot be verified that purchasing power parity is established. In summary, there is a long-term equilibrium relationship between South Korea and developed countries (United States, United Kingdom, Canada), which is consistent with purchasing power parity. Moreover, the price level of developed countries (the United States, the United Kingdom, and Canada) has a greater impact on the exchange rate than that of South Korea; There is no co-integration relationship between developing countries (China) and South Korea. From the perspective of the co-integration equation, the price level of South Korea has a greater impact on the exchange rate than that of China. This may be because the Chinese market has not been fully opened, or because of the influence of the exchange rate system, the Chinese government's intervention in the Chinese exchange rate market is relatively large. When the study time changes, all the data from January 1998 to December 2018 are selected for data analysis. It can be seen that, There is no long-term co-integration relationship between South Korea and the United States, South Korea and China, and purchasing power parity is not established; There is a co-integration relationship between South Korea and the United Kingdom, South Korea and Canada, which is in line with the theory of purchasing power parity, From the perspective of co-integration relationship, the price level of the United Kingdom and Canada has a greater impact on the exchange rate than the price level of South Korea.

한우 암소의 임신기간 및 생시체중에 관한 연구

김대중 전북대학교 일반대학원 2021 국내석사

This study was conducted by analyzing gestation length (GL), birth weight (BW) and reproductive traits (RT) to improve the reproductive efficiency of Hanwoo cows. A total of 5,181 calving records obtained from three Hanwoo breeding farms and one Hanwoo farm with emphasis on reproduction in Jeonbuk Province from 2006 to early 2020 were analyzed. To determine the factors affecting gestation length and calf birth weight, reproductive traits such as year of birth, season of birth, parity of dam, and sex of calf were analyzed by farms. Frequency analysis and descriptive statistics analysis were performed to identify factors affecting cow GL and calf BW. GL and BW data were analyzed by using independent sample t-test and one-way ANOVA and compared by using Scheffe method to determine whether there was a correlation between GL and BW. Multiple regression analysis with IBM SPSS 25.0 statistical program was performed to determine the factors affecting the GL of breeding cow. In this study, the average BW of calves was 28.29kg (SD=4.40), and the average GL was 286.59 days (SD=5.92). There was a statistically significant difference in GL (F=4.431, p=0.005) and BW (F=14.202, p<0.001) for each farm. In this study, the mean GL was 284.84 days (SD=5.87) in 2006-2008, 286.20 days (SD=6.05) in 2009-2011, 287.19 days (SD=5.43) in 2012-2014, 286.97 days (SD=5.72) in 2015-2017, and 286.60 days (SD=6.28) in 2018-2020, respectively. The each mean BW of calves was 26.69kg (SD=4.56) in 2006∼2008, 26.78kg (SD=4.30) in 2009∼2011, 27.40kg (SD=4.26) in 2012∼2014, 28.52kg (SD=3.82) in 2015∼2017 and 29.82kg (SD=4.58) in 2018∼2020. The results showed significant difference between GL (F=17.215, p<0.001) and BW (F=94.232, p<0.001) by YB under the significance level of 5%. In particular, calf BW tended to increase as the YB was more recent. The mean GL by season of birth (SB) was 287.35 days (SD=5.67) in 'spring', 284.74 days (SD=6.01) in 'summer', 287.16 days (SD=5.98) in 'autumn', and 286.81 days (SD=5.68) in winter. In addition, the average BW of calves was 28.83kg (SD=4.36) in 'spring', 27.49kg (SD=4.11) in 'summer', 27.93kg (SD=4.45) in fall, and 28.79kg (SD=4.45) in 'winter'. SD=4.60). Under the significance level of 5%, there were significant differences in GL (F=56.629, p<0.001) and birth weight (F=29.867, p<0.001) by SB. As a result of Scheffe's method for multiple comparisons, the shortest GL was 284.74±6.01 days when the season of birth was summer. And the BW of calves in summer and autumn was smaller than spring and winter. There was a significant difference between GL and BW under the significance level of 5% by the parity of dam (PD) in the gestation length (F=21.261, p<0.001) and birth weight (F=64.323, p<0.001). As a result of Scheffe's method for multiple comparisons, the GL of 1st parity was shorter than 3th parity whereas cows in 7th parity had longer GL than 1st and 2nd parity. Calves from 1st parity dam had the lightest BW. The mean BW of female calves was 27.17kg (SD=4.03) and male calves mean BW was 29.39kg (SD=4.47) and the average GL conceiving female calf was 286.06 days (SD=5.90) while the mean GL conceiving male calf was 287.11days (SD=5.90). Bull calves were carried longer (p<0.05) in utero and its BW was heavier (p<0.05). BW by GL is 25.50kg (SD=5.09) when GL is 270 days or less, 24.79kg (SD=4.59) for 271∼275 days, 25.73kg (SD=3.94) for 276∼280 days, 27.30kg (SD=3.83) for 281∼285 days, 28.67kg (SD=3.96) for 286∼290 days, 30.13kg (SD=4.34) for 291∼295 days, 31.77kg (SD=4.71) for 296∼300 days and 30.77kg (SD=5.75) for above 301 days. Under the significance level of 5%, BW was significantly different (F=118.804, p<0.001) by GL and calf's BW tended to increase as GL increased. The regression coefficients of reproductive traits affecting BW were YB (B=0.295, p<0.001), SB (summer) (B=-0.737, p<0.001), SB (fall) (B=-0.689), p<0.001), PD (B=0.475, p<0.001), SC (female) (B=-2.046, p<0.001), and GL (B=0.220, p<0.001). BW increased by an average of 0.295kg with the passing of one year in the YB and 0.475kg as the PD increased by 1 parity. The average BW of male calves was heavier by 2.046kg than female calves. The average BW increased by 0.22kg as GL was longer by one day. The factor affecting calf's BW was the GL (β=0.296), followed by the YB (β=0.258), SC (β=-0.232), parity (β=0.208), SB (summer) (β=-0.072) and SB (autumn) (β=-0.066). This study indicated that there was a correlation between the GL and BW. Further studies are needed including conception and calving data from other regions for providing criteria for the reproductive capacity of cows, to help in selection super cow. It is necessary to estimate the expected delivery date more accurately, and analyze the calf's BW and weaning weight and consequently it will be helpful to improve reproductive efficiency and prevent reproductive disorders.

Parity of θc particle from QCD sum rule

손영수 연세대학교 대학원 2006 국내석사

We calculate the hadronic properties of the recently discovered penta-quark baryon state, theta_c particle, by using QCD sum rule technique. The charge of theta_c particle is zero, and theta_c is an anti-charmed pentaquark baryon particle, which consists of five quarks. In this dissertation we calculate the parity of theta_c particle by using interpolating field J= [uCu]d(cd). Our result suggests that the parity of theta_c is positive. This is consistent with the previous result of H. Kim, S. H. Lee, Y. Oh using a different interpolating field. 최근에 발견된 theta_c 입자에 대한 특성을 QCD Sum Rule을 이용하여 구해보고자 한다. Theta_c 입자는 charge가 0이고, quark의 수가 5개인 anti-charmed penta-quark baryon 입자이다. QCD Sum rule을 이용하여 계산하는데 있어서 우리는 interpolating field로 J= [uCu]d(cd) 을 사용한다. 그 결과로 우리는 theta_c입자의 parity를 예측해 볼 수 있다. 우리의 결과는 theta_c입자의 parity가 양수(positive)임을 보여준다. 이것은 이미 이전에 다른 interpolating field를 사용하여 theta_c입자의 parity를 구한 H. Kim, S. H. Lee, Y. Oh의 결과와 일치함을 알 수 있다.