Options

How to reproduce Principal Components to original features?

Muhammed_Fatih_Muhammed_Fatih_ Member Posts: 93 Maven
Dear Community, 

is there a possibility in RapidMiner to see which of the original features were selected/transformed to the extracted Principal Components?

Thank you in advance for your help! 

Best regards! 

Fatih   

Answers

  • Options
    ceaperezceaperez Member Posts: 525 Unicorn
    Hi Muhammed, 

    You have three output ports in a PCA operator. 
    The exa port with the results of PCA analysis, the Ori with the original dataset and the pre port with   preprocessing information like eigenvalues and eigenvectors. 

    Which kind of additional information do you need?

    Regards
  • Options
    Muhammed_Fatih_Muhammed_Fatih_ Member Posts: 93 Maven
    Hi @ceaperez

    Thank you for your feedback!

    I want to know which of the original features were considered within the extraction of the respective principal components.

    So, for exmaple: Which original features were used for building PC1?

    Is there a possibility to comprehend the extraction process? 

    Best regards, 

    Fatih
  • Options
    MarcoBarradasMarcoBarradas Administrator, Employee, RapidMiner Certified Analyst, Member Posts: 272 Unicorn
    edited March 2021
    Hi @ceaperez and @mohhamed_syed maybe this link could help you

    Data Science: Concepts and Practice

    https://rapidminer.com/resource/data-science-concepts-practice/

    Page: 470

    Or this video
    https://academy.rapidminer.com/learn/video/workshop-pca
Sign In or Register to comment.