The article uses nonlinear modeling of modern statistical to simulate and analyze the effects of landscape pattern by Natural disturbance and human disturbance.1. It chooses 10 indices as a natural disturbance factors such as size, volume, age, soil thickness, shrub cover, herbaceous cover, average altitude, slope, aspect and slope position to construct of artificial neural network model, multi-model adaptive regression splines, Support vector machine model and project pursuit regression model to analog analyze of fractal index separately.
也说明了研究地的景观斑块大小、景观斑块面积和分维数大部分服从正态分布，部分服从Gama分布。3、运用现代统计学中的非线性建模方法，分析自然干扰和人为干扰对景观格局的影响。1选择面积、蓄积、年龄、土层厚度、灌丛盖度、草本盖度、平均海拔、坡度、坡向和坡位10指标作为自然干扰因子，构建人工神经网络模型、多元自适应回归样条模型、支持向量机模型和投影寻踪回归模型分别模拟分析分维数。
Within the integreted system, the function of the data-basedmanagement sub-system is to input, store, inquire and calculatedata; the function of graphic data management sub-system is toinput, store, inquire, display, output and transform data; thefunction of digital terrain model sub-system is to displaythree dimention graph, calculate slope, aspect and hilling; thefunction of general analysis model sub-system is to make multi-factor statistical analysis such as regression analysis; thefunction of applied model sub-system is to predict the amountof soil erosion, evaluate the land resource and make land useplanning.
属性数据管理子系统具有对属性数据进行输入、存储、检索、统计的功能；图形数据管理子系统具有对图形进行输入、存储、查询、显示、输出及数据格式转换的功能；数字地形模型子系统具有显示三维图形及计算坡度、坡向和阴影的功能；一般分析模型提供了在多因子分析中常用的诸如回归分析等模型；应用模型开发了计算土壤侵蚀量、进行土地资源评价及土地利用规划的模型。
Regression analysis show social goal effect is weakly, however, personal goal and behavior mode effect are significance, self-confident, study orientation and self-confident effect are significance in all subdivision aspect of values, menwhile, study adaptation can also affect values.
4路径分析显示，总体价值观、个人目标、行为方式对学习适应的影响是通过中介变量而间接发生的。直接影响学习适应性的路径主要有三条：学习取向、自制、自信，其中，自制的路径系数最大，说明自制对学习适应的影响力最大。
In order to improve the learning efficiency of the classifier and solve the detecton problem of the image with complicate scenes, a approach which use the linear regression stump as the weak classifier is advanced. The approach weights to the pretiction model from the positive and negative classification, but not to weight it only from the positive aspect.
为了提高分类器的学习效率，解决复杂场景中人体检测的难题，提出了一种利用线性回归末端作为弱分类器的方法，从正、负两个方面对预测模型进行加权，改变了原来的仅从正预测进行加权的方法。
Firstly, we have summarized the research on international and domestic gas price, and at the same time we summed up the methods and models of this forecast. And the several facts which influent the price of gas was carefully analyzed, and abstracted some important factors and then brought forward the model this thesis refers to. Secondly, we discussed theory and arithmetic of partial least squares regression and theory and arithmetic of forecast aspect of BP artificial neural network, thirdly we synthesized the merits of partial least squares regression and theory of non-linear neural network and formed a partial least squares regression and neural network model, using data in natural gas price system in USA market to fit that model, and using acquired data that we have already known to make emulate-forecast and get the error to check over the models accuracy, we use the model to analyze and emulate present natural gas price in China and make a conclusion that Chinese government have made too low gas price.
首先，我们对国际以及国内天然气价格方面的研究做了适当的综述，同时对预测的方法和模型做了一定的归纳和重点分析了影响天然气价格的若干个因素，并提取了认为比较重要的因素作为我们研究的对象，提出本文研究拟构建的模型；其次，研究了偏最小二乘回归方法的原理以及算法和BP神经网络在预测方面的原理和算法，并引入改进的BP算法神经网络模型；再次，我们综合偏最小二乘提取主成分的优点和神经网络非线性的原理构建了偏最小二乘神经网络模型，利用美国这个市场经济下的天然气价格系统中的数据对所构建的模型进行拟合，同时利用已知数据对拟合的模型进行仿真预测，验证此模型的精确性，利用模型对中国目前的天然气价格进行假设仿真，得出近年中国政府的天然气定价过低；最后是结论和建议。
This dissertation also discuss the problem of this aspect. A multidimensional structural regression model for causal inference is established to estimate the population average treatment effect under strongly ignorable treatment assignment.
本文也进行了这方面的探讨，在暴露为强可忽略的假设下，建立了因果推断的多维结构回归模型来估计总体平均因果效应。
Finally, the regression (equations) between the researched soil physical properties and altitude, soil aspect, slope gradient, slope position, soil depth, and ground vegetation coverage were established.
其次，还建立了土壤物理性质与海拔、坡向、坡度、坡位、土壤深度、地表植被覆盖度等之间的回归方程。