CAJ | 학술논문

首先在36种水果中，采用主成分分析的方法选取了菠萝、苹果、橙子、荔枝、木瓜、李子、芒果和樱桃8种主要水果。然后用线性回归和时间序列的方法建立两类模型：第一种，以主要水果的消费量作为因变量，时间作为自变量，建立线性回归模型，经过模型检验、回归系数检验和残差的正态性检验等说明所建的模型是合理的；第二种，用非平稳时间序列的方法建立了时间关于消费量之间的ARIMA（1，1，0）模型，并检验了所建的模型的合理性。另外，得到了主要水果的消费量随着时间而逐年增长的发展趋势。
수선재36충수과중，채용주성분분석적방법선취료파라、평과、등자、려지、목과、리자、망과화앵도8충주요수과。연후용선성회귀화시간서렬적방법건립량류모형：제일충，이주요수과적소비량작위인변량，시간작위자변량，건립선성회귀모형，경과모형검험、회귀계수검험화잔차적정태성검험등설명소건적모형시합리적；제이충，용비평은시간서렬적방법건립료시간관우소비량지간적ARIMA（1，1，0）모형，병검험료소건적모형적합이성。령외，득도료주요수과적소비량수착시간이축년증장적발전추세。
In this paper, firstly we use principal component analysis method to select pineapple, apples,oranges, litchi, pa-paya, plums, mango and cherry eight kinds of major fruit from thirty six kinds of fruit.And then we establish two models with linear regression and time series methods .In the first place, we select the consumption of the main fruit as the dependent var-iable and time as the independent variable, and establish linear regression model.Through the model test, regression coeffi-cient test and residual normality test , it is shown that the model is reasonable.In the second place, using the method of non-stationary time series ,we establish ARIMA (1,1,0) model on the consumption and time , and test the rationality of the model.In addition, we obtain the trend that the consumption of major fruit has been growing with time.