灰色系统模型对水产品产量的实证分析

发布时间：2018-04-05 08:27

本文选题：中国水产品　切入点：高阶GM(n　出处：《大连海洋大学》2015年硕士论文

【摘要】：中国水产品是国民经济的重要组成部分,是直接关系民计民生的重要环节。中国水产品产量一直保持着平稳较快的增长,对此数据进行实证分析具有积极的现实意义。中国水产品产量数据具有指数增长、阶段性、随机性等特征;又由于政治、经济、环境、自然等因素的影响,数据也呈现模糊性、微震荡性,故采用现代时间序列建模方法对数据进行分析具有可行性。这一系列的数据随时间推移既有一定的趋势性又有随机性特征,较为符合灰色系统模型的模型特点。故本文最终选用灰色系统模型进行数据分析。在研究过程中,首先利用传统的GM(1,1)模型对中国水产品产量进行了建模分析,并利用高阶灰色系统模型进行不同阶数建模结果的分析比对,在计算精度与计算成本的综合衡量下,选择了GM(2,1)、GM(3,1)与GM(1,1)对中国水产品产量进行了实证比对分析,得出高阶灰色系统GM(2,1)模型的建模结果较比传统模型精度上有了较大程度的提高,但由于数据的阶段性的影响,高阶灰色系统GM(3,1)模型的建模结果则误差较大,不宜进行实证分析。同时,在中国水产品产量所包含的海洋捕捞、远洋渔业、海水养殖、淡水捕捞、淡水养殖产量数据中,为了分析各产量之间的潜在关系,利用灰关联分析按序列的关联关系将中国水产品产量分为两组,并利用多变量灰色系统模型对分组后的两组数据分别进行建模分析,误差计算结果显示了建模结果的有效性。通过以上实证分析可知,灰色系统理论中的高阶灰色系统模型和多变量灰色系统模型对中国水产品产量数据进行分析能够收到良好的效果,但仍存有数据特殊性的限制;第二,利用高阶灰色系统模型对呈指数型增长且伴有随机性的数据进行建模,以及利用多变量灰色系统模型对有内在关联的数据进行分组建模具有推广意义;第三,从追求单数据的建模精度到利用多维数据关系进行建模的思路在当今大数据的背景下,具有现实意义。这为相关部门决策与分析生产与供给、消费与需求、进出口、市场流通和价格等问题提供了数据分析基础,具有借鉴意义。
[Abstract]:Aquatic products is an important part of the national economy and an important link directly related to people's livelihood.The production of aquatic products in China has been increasing steadily and rapidly.China's aquatic product output data are characterized by exponential growth, stages, randomness, and so on. Due to the influence of political, economic, environmental, natural and other factors, the data also show fuzziness and micro-volatility.Therefore, it is feasible to use the modern time series modeling method to analyze the data.This series of data not only has some tendency and randomness with time, but also accords with the model characteristic of grey system model.So this paper chooses the grey system model to analyze the data.In the course of the research, the traditional GM1 / 1) model is used to model and analyze the aquatic product production in China, and the high order grey system model is used to analyze and compare the different order modeling results.In this paper, we have selected GMGM2GM1) to compare and analyze the aquatic product production in China. It is concluded that the modeling result of the high-order grey system GMGM2GM1) model is higher than that of the traditional model, but because of the influence of the stage of the data, the model is more accurate than that of the traditional model, but because of the effect of the stage of the data,High order grey system GM3 + 1) model is not suitable for empirical analysis because of its large error.At the same time, in order to analyze the potential relationship between the various yields, in the data on marine fishing, ocean fisheries, mariculture, freshwater fishing and freshwater aquaculture, which are included in China's aquatic product production,The production of aquatic products in China is divided into two groups by using the grey correlation analysis according to the correlation relation of the sequence, and the two groups of data after grouping are modeled and analyzed by using the multivariable grey system model. The result of error calculation shows the validity of the modeling results.Through the above empirical analysis, we can see that the high-order grey system model and the multi-variable grey system model in the grey system theory can achieve good results in the analysis of aquatic product production data in China, but there are still some limitations of the data particularity.Using the high-order grey system model to model the exponential growth data with randomness, and using the multi-variable grey system model to group the data with internal correlation is of great significance.From pursuing the modeling accuracy of single data to modeling by using multidimensional data relationship, it is of practical significance under the background of big data today.It provides a data analysis basis for the decision and analysis of production and supply, consumption and demand, import and export, market circulation and price.
【学位授予单位】：大连海洋大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F326.4;N941.5

【相似文献】