+ Reply to Thread
Results 1 to 6 of 6

Thread: "Linear at the Parameters"

  1. #1
    Points: 3,954, Level: 40
    Level completed: 2%, Points required for next Level: 196

    Posts
    9
    Thanks
    2
    Thanked 0 Times in 0 Posts

    SOLVED: "Linear at the Parameters"




    I'm having a hard time with this concept. I've read as many different sources on this terminology as I can find, but it seems that most of them summarily describe it the same way without really expounding in a manner that helps me build a conceptual understanding. I thought I might have finally started to grasp this notion, but then I fell flat on my face with the first homework question.

    The model provided is: Y = 10,000 + 150*X1 + 25*X1^2 + 60*X2
    The question is: Is this a “linear” regression model, why or why not?

    Based on the concept as I derived it from reading, I thought it was not a linear model. Based on subsequent questions in the class work, I'm pretty darn sure that it is supposed to be linear.

    Could someone detangle my brain for me?
    Last edited by JTWood; 10-18-2014 at 08:59 AM.

  2. #2
    Super Moderator
    Points: 13,151, Level: 74
    Level completed: 76%, Points required for next Level: 99
    Dragan's Avatar
    Location
    Illinois, US
    Posts
    2,014
    Thanks
    0
    Thanked 223 Times in 192 Posts

    Re: "Linear at the Parameters"

    I think what you need to do is understand the difference between "Linearity in the Variables" versus "Linearity in the Parameters". In your example that you provided above Y is not a "linear function" because the variable X1 appears as a power or index of 2.

    That said, to be a "Linear function of the Parameters" the model Y = b0 +b1*x^2 is a "Linear Regression Model" but Y = b0 + Sqrt[b1]*X is not - i.e. the latter in an example of a non-linear (in parameters) regression model.

    Hope this helps.

  3. The Following User Says Thank You to Dragan For This Useful Post:

    JTWood (10-18-2014)

  4. #3
    Points: 3,954, Level: 40
    Level completed: 2%, Points required for next Level: 196

    Posts
    9
    Thanks
    2
    Thanked 0 Times in 0 Posts

    Re: "Linear at the Parameters"

    So "linear at the parameters" is a condition where the parameter itself (b1, b2, etc) is linear regardless of what is happening to the variables?
    As I understand it, a model/function will be considered linear if the parameters are linear and even when the visual depiction of the model is not linear itself.
    Putting the two together, it would be appropriate to say that the original model is linear because it is linear at the parameters (10,000/150/25/60).

    Does that sound right?

  5. #4
    Super Moderator
    Points: 13,151, Level: 74
    Level completed: 76%, Points required for next Level: 99
    Dragan's Avatar
    Location
    Illinois, US
    Posts
    2,014
    Thanks
    0
    Thanked 223 Times in 192 Posts

    Re: "Linear at the Parameters"

    Quote Originally Posted by JTWood View Post

    Does that sound right?
    Yes, sounds good to me.

  6. #5
    Points: 3,954, Level: 40
    Level completed: 2%, Points required for next Level: 196

    Posts
    9
    Thanks
    2
    Thanked 0 Times in 0 Posts

    Re: "Linear at the Parameters"

    Thanks! I feel better about this now.

  7. #6
    Super Moderator
    Points: 13,151, Level: 74
    Level completed: 76%, Points required for next Level: 99
    Dragan's Avatar
    Location
    Illinois, US
    Posts
    2,014
    Thanks
    0
    Thanked 223 Times in 192 Posts

    Re: "Linear at the Parameters"


    Okee Dokee....

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats