问题描述
我做了2个实验
第一个实验:在训练实验中仅将PCA用作数值数据，当我创建预测性实验并预测一个记录时，对于主成分，它返回全零
1st experiment: uses PCA for only numerical data in the training experiment, when i created the predicted experiment, and predicted for one record, it returned all zero for the the Principal components
第二个实验:我在训练实验中同时使用了PCA作为字符串和数字数据，当我创建预测性实验并预测了一条记录时，它返回了主要成分的值
2nd experiment: I used PCA for both string and numerical data in training experiment, when i created the predicted experiment, and predicted for one record, it returned values for the the Principal components
为什么第一个实验返回的全为零?
why 1st experiment is returning all zero?
推荐答案
我怀疑PCA无法降低实验1中数值变量的维数.
I suspect the PCA is not able to reduce the dimensionality of your numerical variables in your experiment-1.
https://georgemdallas.wordpress.com/2013/10/30/principal-component-analysis-4-dummies-eigenvectors-eigenvalues-and-dimension-reduction/
https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/principal-component-analysis
此致，
Jaya
Regards,
Jaya
这篇关于为什么PCA转换对所有主要成分都返回零?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！