CAJ | 학술논문

根据长白山地区白河林业局的772块固定标准地调查数据，以及最小二乘法（ OLS），建立逻辑斯蒂（Logistic）模型来预估该局现时状态有林地的红松分布概率，并采用泊松（Poisson）和负二项分布（Negative binomi-al，NB）模型预估该局现时状态有林地红松的分布数量，并对模型进行了拟合效果评价及独立性检验。结果表明， Logistic模型与数据的拟合效果很好，独立性检验中预测精度也在可接受的范围内，可以利用Logistic模型来预估该局有林地红松的分布概率。与Poisson模型相比，负二项分布模型能够解决因变量的不均匀分布（即过度散布）的问题，因此能够更好地拟合数据。但在独立性检验中，Poisson模型和NB模型的预测精度接近，且都在可接受的范围内，两个模型均可以用来预估该局有林地红松的分布数量。
근거장백산지구백하임업국적772괴고정표준지조사수거，이급최소이승법（ OLS），건립라집사체（Logistic）모형래예고해국현시상태유임지적홍송분포개솔，병채용박송（Poisson）화부이항분포（Negative binomi-al，NB）모형예고해국현시상태유임지홍송적분포수량，병대모형진행료의합효과평개급독립성검험。결과표명， Logistic모형여수거적의합효과흔호，독립성검험중예측정도야재가접수적범위내，가이이용Logistic모형래예고해국유임지홍송적분포개솔。여Poisson모형상비，부이항분포모형능구해결인변량적불균균분포（즉과도산포）적문제，인차능구경호지의합수거。단재독립성검험중，Poisson모형화NB모형적예측정도접근，차도재가접수적범위내，량개모형균가이용래예고해국유임지홍송적분포수량。
With the survey of 720 fixed plots in Baihe Forestry Bureau of Changbai Mountains , we established logistic model to estimate probability of occurrence of current Korean pine in forest land of Baihe Forestry Bureau , and established Poisson model and negative binomial model to estimate number of Korean pine by using ordinary least square .Goodness of fit tests and independence tests were implemented for all models .The logistic model fitted the data well and had an acceptable pre-diction accuracy , which indicated the usability of logistic model for estimating the probability of occurrence of Korean pine . Negative binomial model fitted the data better than Poisson model due to its ability of solving uneven distribution of the de -pendent variables (over dispersion).However, for independence tests, the Poisson and NB model had similar and accept-able prediction accuracies , which indicated that two models could be used to estimate distribution of the number of Korean pine.