LAND-USE ANALYSIS PERFORMANCE IN PRECISION AGRICULTURE: EVOLUTION ALGORITHM OF DATA MINING

摘要

With the rapid development of economy, Chinese rural land-use is experiencing tremendous change, and the constant retreat of the land-use area causes a huge reaction recently. In this paper, we take the change of crops area and the reasons why changes triggered as the subject of our study. We use the technology of Remote Sensing (RS) to gain data of crops area of Ningbo in the period from 1986 to 2005, when China was conducting the reform and opening up and reconstruction of agriculture, which caused a lot of change in crops area. We adopt Data Mining (DM) to gain knowledge and analyze this phenomenon, as well as predicate the tendency of the crops area in the future. In this paper, the evolution algorithm of data mining is used to model the changes and the tendency of crops coverage as time goes by. We take into account four factors: State Management (SM), Local Management (LM), Market (MK) and Weather (WH). In our research, we found firstly the factor of "SM" is a dominant factor but its influence will decline in the future, secondly the factor of "MK" is a rising and important factor, thirdly the factor of "LM" is in the rising stage but it wouldn't have more share than the factor of "MK", and finally "WH" factor which accounts for a large proportion in the past and now is declining to the lightest one. According to our experiment, it is predicated that Chinese crops area downward trend will decelerate, on the contrary it will have a continuous rising until it gets to the optimization conditions and then realize the harmonious use of the land.