2/24/2017

try several models

Today, I tried several models. They include Cubist, Random Forest and SVM. But after trials and errors, I could not find that they have good modeling results.

Summary:

 SVM:
 Random Forest:
Cubist:
All of their R2 are lower than 0.5, which show bad modeling results.

Yesterday's ANN flow chart:


I find that there are also some packages of ANN and other models in Matlab. 
Next week, I plan to look into them to find if they can be helpful for my research.



Posted by at

4 comments:

  1. siddharth misraSaturday, February 25, 2017 at 1:03:00 AM CST

    Stick to R language. The goal is to extend R to log interpretation. R is open source compared to Matlab.

    ReplyDelete
  2. siddharth misraSaturday, February 25, 2017 at 1:07:00 AM CST

    you need to look into these pending tasks

    2. get an account in OSCER to do the lengthy computation.

    3. show me in slides all the steps for a known data..apply all these methods to real data... we a want to see the steps/methods work for real data..

    4. look into one formation..and apply the method to data from one and nearby formations rather than using the entire depth range

    ReplyDelete
    Replies
    1. UnknownSaturday, February 25, 2017 at 11:57:00 AM CST

      For 2, only ANN has lengthy computation, which also shows a bad modeling result. So for now, there is no need to use OSCER.
      For 3, I will show you the slides. I have applied these methods to real data. The results are shown in the blog.
      For 4, I have looked into one formation, the results are not good, either.

      Delete

Subscribe to: Post Comments (Atom)