时拓

这些结果在应用Hopfield网络求解优化问题和联想记忆时拓广了神经元激活函数的选择范围。

These results extend the choice range of activation function of neuron when using Hopfield net to solution of optimization problem or to associative memory .

通过二叉树、多叉树和链表相结合的数据结构来记录实体的特征造型历史，保存特征重构前后拓扑元素之间的关系，同时为特征重构时所需要拓扑面和边设计了查询算法。

The binary tree , multi tree and list are used to record the modeling history and the relationship between the topological entities before and after feature-rebuilding . A researching approach to the topological faces and edges which the feature-rebuilding needs is designed .

偏移时，通过延拓算子查表法，利用McClellan变换实现三维波场的向下延拓。

1-D Extrapolation operator with McClellan transformation can achieve 3-D table-driven migration .

同时提出了基于该滤波器进行数据重构时的边界延拓方法。

Boundary extension method for data reconstruction based on this filter is also proposed .

给出了数字图像的小波变换及边界延拓的步骤。并且，提升方法对宿主图像的尺寸没有特别的要求，而且也没有一代小波变换时的边界延拓问题。

At the same time , the steps of wavelet transform and boundary extension for digital image are given . In addition , lifting method has no requirements for image size .