码迷,mamicode.com
首页 > 其他好文 > 详细

Strassen algorithm(O(n^lg7))

时间:2017-06-02 23:08:53      阅读:219      评论:0      收藏:0      [点我收藏+]

标签:src   rms   each   ring   amp   block   cab   exce   isp   

Let AB be two square matrices over a ring R. We want to calculate the matrix product C as

{\displaystyle \mathbf {C} =\mathbf {A} \mathbf {B} \qquad \mathbf {A} ,\mathbf {B} ,\mathbf {C} \in R^{2^{n}\times 2^{n}}}技术分享

If the matrices AB are not of type 2n × 2n we fill the missing rows and columns with zeros.

We partition AB and C into equally sized block matrices

{\displaystyle \mathbf {A} ={\begin{bmatrix}\mathbf {A} _{1,1}&\mathbf {A} _{1,2}\\\mathbf {A} _{2,1}&\mathbf {A} _{2,2}\end{bmatrix}}{\mbox{ , }}\mathbf {B} ={\begin{bmatrix}\mathbf {B} _{1,1}&\mathbf {B} _{1,2}\\\mathbf {B} _{2,1}&\mathbf {B} _{2,2}\end{bmatrix}}{\mbox{ , }}\mathbf {C} ={\begin{bmatrix}\mathbf {C} _{1,1}&\mathbf {C} _{1,2}\\\mathbf {C} _{2,1}&\mathbf {C} _{2,2}\end{bmatrix}}}技术分享

with

{\displaystyle \mathbf {A} _{i,j},\mathbf {B} _{i,j},\mathbf {C} _{i,j}\in R^{2^{n-1}\times 2^{n-1}}}技术分享

then

{\displaystyle \mathbf {C} _{1,1}=\mathbf {A} _{1,1}\mathbf {B} _{1,1}+\mathbf {A} _{1,2}\mathbf {B} _{2,1}}技术分享
{\displaystyle \mathbf {C} _{1,2}=\mathbf {A} _{1,1}\mathbf {B} _{1,2}+\mathbf {A} _{1,2}\mathbf {B} _{2,2}}技术分享
{\displaystyle \mathbf {C} _{2,1}=\mathbf {A} _{2,1}\mathbf {B} _{1,1}+\mathbf {A} _{2,2}\mathbf {B} _{2,1}}技术分享
{\displaystyle \mathbf {C} _{2,2}=\mathbf {A} _{2,1}\mathbf {B} _{1,2}+\mathbf {A} _{2,2}\mathbf {B} _{2,2}}技术分享

With this construction we have not reduced the number of multiplications. We still need 8 multiplications to calculate the Ci,j matrices, the same number of multiplications we need when using standard matrix multiplication.

Now comes the important part. We define new matrices

{\displaystyle \mathbf {M} _{1}:=(\mathbf {A} _{1,1}+\mathbf {A} _{2,2})(\mathbf {B} _{1,1}+\mathbf {B} _{2,2})}技术分享
{\displaystyle \mathbf {M} _{2}:=(\mathbf {A} _{2,1}+\mathbf {A} _{2,2})\mathbf {B} _{1,1}}技术分享
{\displaystyle \mathbf {M} _{3}:=\mathbf {A} _{1,1}(\mathbf {B} _{1,2}-\mathbf {B} _{2,2})}技术分享
{\displaystyle \mathbf {M} _{4}:=\mathbf {A} _{2,2}(\mathbf {B} _{2,1}-\mathbf {B} _{1,1})}技术分享
{\displaystyle \mathbf {M} _{5}:=(\mathbf {A} _{1,1}+\mathbf {A} _{1,2})\mathbf {B} _{2,2}}技术分享
{\displaystyle \mathbf {M} _{6}:=(\mathbf {A} _{2,1}-\mathbf {A} _{1,1})(\mathbf {B} _{1,1}+\mathbf {B} _{1,2})}技术分享
{\displaystyle \mathbf {M} _{7}:=(\mathbf {A} _{1,2}-\mathbf {A} _{2,2})(\mathbf {B} _{2,1}+\mathbf {B} _{2,2})}技术分享

only using 7 multiplications (one for each Mk) instead of 8. We may now express the Ci,j in terms of Mk, like this:

{\displaystyle \mathbf {C} _{1,1}=\mathbf {M} _{1}+\mathbf {M} _{4}-\mathbf {M} _{5}+\mathbf {M} _{7}}技术分享
{\displaystyle \mathbf {C} _{1,2}=\mathbf {M} _{3}+\mathbf {M} _{5}}技术分享
{\displaystyle \mathbf {C} _{2,1}=\mathbf {M} _{2}+\mathbf {M} _{4}}技术分享
{\displaystyle \mathbf {C} _{2,2}=\mathbf {M} _{1}-\mathbf {M} _{2}+\mathbf {M} _{3}+\mathbf {M} _{6}}技术分享

We iterate this division process n times (recursively) until the submatrices degenerate into numbers (elements of the ring R). The resulting product will be padded with zeroes just like A and B, and should be stripped of the corresponding rows and columns.

Practical implementations of Strassen‘s algorithm switch to standard methods of matrix multiplication for small enough submatrices, for which those algorithms are more efficient. The particular crossover point for which Strassen‘s algorithm is more efficient depends on the specific implementation and hardware. Earlier authors had estimated that Strassen‘s algorithm is faster for matrices with widths from 32 to 128 for optimized implementations. However, it has been observed that this crossover point has been increasing in recent years, and a 2010 study found that even a single step of Strassen‘s algorithm is often not beneficial on current architectures, compared to a highly optimized traditional multiplication, until matrix sizes exceed 1000 or more, and even for matrix sizes of several thousand the benefit is typically marginal at best (around 10% or less).

 from Wikipedia

--------------------------------------------------------------------------------------------------------------------------------------------------------------

it substitude the 8th recursive invocation(multiplication) by the liner combination of the submatrices above(cause A4,4 and B4,4 has been used before).
like a*(b+c) can have less steps than a*b+a*c,it uses liner combination to simplify the tranditional multiplicate way.

 

Strassen algorithm(O(n^lg7))

标签:src   rms   each   ring   amp   block   cab   exce   isp   

原文地址:http://www.cnblogs.com/wujunde/p/6935348.html

(0)
(0)
   
举报
评论 一句话评论(0
登录后才能评论!
© 2014 mamicode.com 版权所有  联系我们:gaon5@hotmail.com
迷上了代码!