乘法

A new block cipher is designed where the diffusion of matrix multiplication is combined with the confusion of specially designed matrix operation to encrypt information.

设计了一种新的分组密码算法，该算法利用矩阵乘法的扩散作用与专门设计的一种矩阵运算的混乱作用实现对信息的加密。

Since the information for the generalized Jordan canonical blocks in the generalized Jordan canonical form can not be easily obtained from the characteristic polynomial and the commutative law does not hold for matrix multiplication,we compute the common form of the power of the generalized Jordan canonical form and get a concise expression for the power of the defective matrix.

考虑到不易从它的特征多项式获得构成广义若当标准型的阶数更小的广义若当块的有关信息，针对矩阵乘法不满足交换率，本文从计算广义若当标准型的幂的一般形式出发，获得实数域上亏损矩阵的幂的一个简洁表示。

Keep in mind that, as is the case with matrices, quaternion multiplication is not commutative!

紧记，矩阵在此种情况下，四元数的乘法是不可进行交换的。

You might also know (or remember from your math lessons) that matrix multiplication is not commutative.

你可能也知道的矩阵乘法是不可交换的。

Validating experiment was carried out on IKONOS2 images, and showed that resultant regression coefficients were more acceptable for their physical meanings, and that the discribed method performed comparable with spectral response function or least squares based methods.

通过IKONOS-2全色与多光谱图像的融合实验，结果表明：该方法所求得的回归系数具有明确物理意义，符合实际情况，并且与光谱响应函数法、最小二乘法相比，其融合质量基本保持一致，并且由于该方法不需要先验知识，故其实用性较强。

To a certain extent, it has overcome proportionality factor influence and the displacement difference of the Deng interrelatedness analysis method. The experiment proves that in the examination peripheral point's continuity, the complete aspect has certain improvement compared to the Deng interrelatedness model examination.

根据邓氏关联度分析方法建立的改进模型，结合仿射变换和最小二乘法直线拟合，一定程度上克服了邓氏关联度分析方法的比例因子影响和位移差，实验证明在检测边缘点的连续性、完整性方面比用邓氏关联度模型检测的边缘有一定的改进。

The multiplier forms the partial products of the complex multiplication and produces a 32 bit signed fraction result.

The乘数形式部分产品的生产复杂的乘法和32位有符号的部分结果。

In comparison with traditional pipeline FFT, the complex multiplication was implemented by shifting and addition operations in order to eliminate multiplication unit and coefficient ROM.

与传统的流水线FFT比较，该模块中的复数乘法运算全部采用移位相加操作完成，因而消除了乘法器及旋转因子ROM的使用，降低了功耗。

Signal processing approach presented in this paper based on neural network algorithm (spectral analysis, Random noise filtering) does not involve the complex multiplication and addition operations. Its computing accuracy is high. It is particularly suitable for the implementation of software and hardware based on DSP chip.

所提出的基于神经网络算法的信号处理方法(频谱分析、随机噪声滤波)不涉及复数的乘法运算和复数的加法运算，计算精度高，特别适合基于DSP芯片的软、硬件实现。

Using bi-imaginary number representation, complex multiplication is simplified into two multiplication, instead of four ones in the previous method.

为克服空间带宽积的限制，提出了利用有限的外积乘法单元，通过矩阵分块和矢量循环流动进行任意维数矩阵与矢量运算算法，处理器利用率高。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。