投影寻踪模型的改进及其在生态农业建设综合评价中的应用

为提高生态农业建设综合评价模型的分辨率，提出了一种有效的和通用的模型——投影寻踪模型(PP)。模型中密度窗宽R是求解局部密度的窗口半径，是由样本数据本身特性确定的局部宽度参数，主要通过试算或经验来确定，缺乏理论根据，该文对模型中密度窗宽R进行了理论改进，推导得出了计算的经验公式，使模型更具有科学性和稳定性。采用基于实码的加速遗传算法寻找最优的投影方向，同时用投影方向信息研究了各因子对生态农业建设综合评价的影响水平，取得了符合客观实际的分类结果，为生态农业合理建设提供了决策依据。

Improvement of projection pursuit model and its application in integrated evaluation of eco-agricultural construction

In order to raise the distinguishment of the model of integrated evaluation of eco-agricultural construction, an effective and general model-PP(Projection Pursuit) model was presented. The density window breadth R is the window radius that solves the partial density, it is determined by the characteristic of the sample datum, and it is mainly determined by trial calculation or experience, which lacks theoretical basis. This article improves the density window breadth R of the model in theory, it deduces and acquires the empirical formula of calculation, making the model more scientific and stable. Real Coding based on Accelerating Genetic Algorithm was adopted to find the best projective direction, at the same time, the datum of the best projective direction was used to research the level of the influence of each factor to eco-agricultural construction. The classification results which accord with the fact were gained, which provide the decision proof for constructing eco-agricultural reasonably.