The query in Example 6-7 returns the three attributes that have the greatest impact on the top PCA projection for customer 101501. The FEATURE_DETAILS
function applies a Singular Value Decomposition model named svd_sh_sample
to the data selected from svd_sh_sample_build_num
.
SELECT FEATURE_DETAILS(svd_sh_sample, 1, 3 USING *) proj1det FROM svd_sh_sample_build_num WHERE CUST_ID = 101501; PROJ1DET -------------------------------------------------------------------------------- <Details algorithm="Singular Value Decomposition" feature="1"> <Attribute name="HOME_THEATER_PACKAGE" actualValue="1" weight=".352" rank="1"/> <Attribute name="Y_BOX_GAMES" actualValue="0" weight=".249" rank="2"/> <Attribute name="AGE" actualValue="41" weight=".063" rank="3"/> </Details>