Hi,I have taken up Attrition analysis as the proj...

Discussion in 'Big Data and Analytics' started by Poornima M_2, Oct 12, 2017.

  1. Poornima M_2

    Poornima M_2 Member

    Joined:
    May 15, 2017
    Messages:
    10
    Likes Received:
    0
    Hi,
    I have taken up Attrition analysis as the project.As per the guidelines,i have done Descriptive analysis and regression TEST for the same
    Now,i am having a doubt.
    It is mentioned to output those observation in a dataset whose value is more than the cutoff value.
    Now how can we compute the cutoff value and with what value must this be compared to.?I have performed a ttest and took the pr value as the cutoff value
    Pls confirm if this is correct.

    Thanks,
    Poornima
     
    #1
  2. Vikram_13

    Vikram_13 Member
    Simplilearn Support

    Joined:
    Nov 25, 2015
    Messages:
    11
    Likes Received:
    0
    First check that there will at-least one variable which will be significant , if none of the variable is significant , it does not make sense to build the model in that data .

    You will set up a threshold value like may be prob val greater than 50 percent all customers will churn . Identify all those customers and prep new set from them.
     
    #2

Share This Page