历史查询   



 没有找到相关例句 
 In this paper, 6 phrases of satellite images include Corona?TM?SPOT?ETM+ data which represent the information of Changshu of 1966,1984,1992,1999,2000,2001 respectively were dealt with multi band composition, image enhancement, precision emendation, interpretation index establishment, computer assistant visual interpretation. Vector maps of expanding towns of Changshu City from 1966 to 2001 are made by interpreting images above. The statistical result shows that the area of city is 6.4 times bigger in 2001... In this paper, 6 phrases of satellite images include Corona?TM?SPOT?ETM+ data which represent the information of Changshu of 1966,1984,1992,1999,2000,2001 respectively were dealt with multi band composition, image enhancement, precision emendation, interpretation index establishment, computer assistant visual interpretation. Vector maps of expanding towns of Changshu City from 1966 to 2001 are made by interpreting images above. The statistical result shows that the area of city is 6.4 times bigger in 2001 than in 1966. From 1984 to 2001, it becomes bigger and bigger of average area of town of every person for 24 towns. In expending space, city expending is affected by geography, physiognomy condition, economic level and policy decision making. Extending directions of most of towns are accordant with the main rivers and major roads. There is significant difference in amount and relative rate in the extending among 24 towns, and there is geographic disciplinarian in the distribution of towns at different expending levels. Most of towns that have large expanding rate have convenient communication, but towns that have low expanding rates are far from major roads and rivers. According to many years' statistics of city expanding and historical social and economic data, it is found that city expanding is related to rural urban migration, GDP and original dimension evidently, while is not to total population, total families, total industry production and total employee. Through the statistic, we found that there is significant difference in the values of GDP of per urban area among 24 towns.  利用常熟市不同时期 (1 966、1 984、1 992、1 999、2 0 0 0、2 0 0 1年 )、不同类型卫星遥感影像 (Corona、TM、SPOT、ETM +)进行多波段合成 ,图像增强处理与精校正 ,建立不同影像城镇用地解译标志 ,通过人机交互判读 ,获得不同时期的各乡镇城镇用地矢量图的方法。统计出 35年来常熟市城镇建设用地扩大近 6 .4倍。城镇扩展空间上受到地貌条件和城市经济水平以及政策决策影响 ,城镇多数沿河流或交通线方向扩展 ,各城镇间扩展数量和相对速率有明显差异 ,且存在一定地理分布规律。结合社会经济的统计资料 ,分析发现城镇扩展与不断增加的城镇人口、GDP极显著相关 ,各乡镇单位城镇面积创造的GDP有较大的差异。  The hidden unit performance related spaces, i e representation space and error space, target space and expended space, are analyzed according to the model of least squares approximation feedforward neural networks, and further more, the criteria and the evaluation methods for the performance of hidden units are proposed It is revealed that the efficient component of output vector of a hidden unit should be lying in error space, avoiding expended space and closing... The hidden unit performance related spaces, i e representation space and error space, target space and expended space, are analyzed according to the model of least squares approximation feedforward neural networks, and further more, the criteria and the evaluation methods for the performance of hidden units are proposed It is revealed that the efficient component of output vector of a hidden unit should be lying in error space, avoiding expended space and closing the direction of maximum energy, which are independent of the nonlinear functions used by the units, and it is permitted that different units have different activation functions The quality factor of the hidden layer, efficient coefficient of the hidden layer, the redundancy of hidden units and the evaluation factor of the hidden layer for total evaluation, are proposed for the performance evaluation of the hidden layer  针对最佳平方逼近三层前馈神经网络模型 ,分析了与隐层单元性能相关的表示空间与误差空间、目标空间与耗损空间的作用 ,提出了按网络生长方式构建隐层时隐单元选择准则和评价方法 研究结果表明 :隐单元选取策略应遵循其输出向量有效分量位于误差空间、回避耗损空间和尽可能趋向于极大能量方向的原则 ,这一结果与隐单元采用什么激发函数无关 ,也允许各隐单元采用不同激发函数 网络的隐层性能评价可以通过隐层品质因子、隐层有效系数、隐单元剩余度来进行 ,而总体结果可采用隐层评价因子进行评测  The error compensation performance of the hidden unit was discussed according to the model of leastsquares approximation feedforward neural networks using the method of subspace analysis,and further more,a hidden layer evaluation method was proposed.It was revealed that the efficient component of output vector of a hidden unit should be lying in error space,avoiding expend space and closing an energy space,which were independent of the nonlinear functions used by the hidden units,and it was permitted... The error compensation performance of the hidden unit was discussed according to the model of leastsquares approximation feedforward neural networks using the method of subspace analysis,and further more,a hidden layer evaluation method was proposed.It was revealed that the efficient component of output vector of a hidden unit should be lying in error space,avoiding expend space and closing an energy space,which were independent of the nonlinear functions used by the hidden units,and it was permitted that different hidden units have different activation functions.The quality factor of hidden layer,efficient coefficient of hidden layer,the redundancy of hidden units and the evaluation factor of hidden layer,for total evaluation,were proposed to evaluate the performance of the hidden layer.The rationality and validity of proposed method is validated by the results of evaluation experiment.1fig.,1tab.,11refs.  针对最佳平方逼近3层前馈神经网络模型,采用子空间分析方法,讨论了隐单元的误差补偿性能,提出了隐层评测方法.研究结果表明隐单元选取策略应遵循其输出向量有效分量位于误差空间、回避耗损空间和尽可能靠近某一能量空间的原则,这一结果与隐单元采用什么激发函数无关,也允许各隐单元采用不同激发函数.网络的隐层性能评价可以通过隐层品质因子、隐层有效系数、隐单元剩余度来进行,而总体结果可采用隐层评价因子进行评测.评测实验表明,所提出的隐层评测方法是合理有效的.图1,表1,参11.  
相关查询  



