Larger-Scale Problems
Larger-Scale Problems
复制标题
更大规模的问题
DOI：
10.1007/978-0-387-78723-7_14
复制
发表时间：
2008
期刊：
影响因子：
--
通讯作者：
M. Bartholomew
中科院分区：
文献类型：
--
作者：
M. Bartholomew
到目前为止所考虑的大多数例子只涉及几个变量。实际的优化问题往往涉及非常多的未知数，并且随着变量数量的增加，对于小问题表现良好的方法可能变得不那么有效。现在我们研究一下当n变大时，第7-11章中描述的方法的行为。我们用于比较的问题是第4章4.3节中的OC 1（n）和OC 2（n）。表14 - 1显示了在不同的n值下，求解问题OC 1（n）所需的迭代次数和函数调用次数。这些数字与高精度收敛测试（4.3. 2）。n= 50 n= 100 n= 200
Most of the examples considered so far have involved only a few variables. Practical optimization problems often deal with very many unknowns and methods which perform well for small problems may become less efficient as the number of variables increases. We now investigate the behaviour of the methods described in Chapters 7–11 as n becomes larger. The problems we use for our comparison are OC1 (n) and OC2 (n) from Section 4.3 of Chapter 4. Table 14.1 shows numbers of iterations and function calls needed to solve Problem OC1 (n) for various values of n. The figures relate to the high-accuracy convergence test (4.3. 2). n= 50 n= 100 n= 200