A PRODUCT OF DISCRIMINATIVE LEARNING MAY LEAD TO FEMALE PREFERENCES FOR ELABORATE MALES
A PRODUCT OF DISCRIMINATIVE LEARNING MAY LEAD TO FEMALE PREFERENCES FOR ELABORATE MALES
复制标题
歧视性学习的产物可能会导致女性偏爱精致的男性
DOI：
--
复制
发表时间：
1993
期刊：
影响因子：
--
通讯作者：
Ronald G. Weisman
中科院分区：
文献类型：
--
作者：
Daniel M. Weary;Tim Guilford;Ronald G. Weisman
得到的矩阵R(=Gd，d，T)的第四行和第四列的所有元素将为零，其余元素将为非负。此外，可以很容易地证明，只要G是半正定的，R就是半正定的，从而是有效的方差-协方差矩阵(Morison，1976)。在适当地重新定义P“之后，可以使用(AI)进一步分解由R的前三行和前三列组成的3×3矩阵。为了确定通常是否可能满足条件(A3)，我随机生成了4x4相关矩阵，相关系数从0到1之间的均匀分布中选择。在排除非正半定的矩阵后，50例中，除I外，所有的矩阵都有可能重新排列以满足(A3)。在一种例外情况下，在不能满足(A3)的几个人工例子中，总是可以找到另一种方法来表示形式(4)的矩阵，其中dj的所有元素被约束为非负的(Fry，unpubl)。结果)。
All elements of the fourth row and column of the resulting matrix R (= G d,d, T)will be zero, and the remaining elements will be nonnegative. In addition, it can easily be verified that R will be positive semidefinite, and thus will be a valid variance-covariance matrix (Morrison, 1976), as long as G is positive semidefinite. The 3 x 3 matrix consisting of the first three rows and columns of R can then be further decomposed using (A I), after redefining the P" appropriately. To determine whether it would usually be possible to meet the conditions (A3), I randomly generated 4 x 4 correlation matrices, with the correlations chosen from a uniform distribution between zero and one. After excluding matrices that were not positive semidefinite, it was possible to rearrange the matrix to satisfy (A3) in all but I of 50 cases. In the one exceptional case, and in several artificial examples in which (A3) could not be satisfied, it was always possible to find an alternative method for expressing the matrix in the form (4) with all elements of the d j constrained to be nonnegative (Fry, unpubl. results).
自动成形和操作键啄的泛化峰值偏移。
影响因子：
2.7
作者：
Weiss,SJ;Weissman,RD
通讯作者：
Weissman,RD