Skip to main content

Author Topic: Regression Trees

r
  • *****
  • Hero Member
  • Posts: 575
Regression Trees
OP: June 05, 2014, 11:00:00 PM
Has anyone here put LC data through Regression Trees?  I just came across this stuff on a work assignment and they used it on a portfolio of credit to find advance relationships that would be too difficult to detect without big data technology.  I'm considering signing up for the free trial and running LC data through it to see what it shows but was curious if anyone here had done it or had experience with it. And yes, I'm looking at you Fred, AnilG, and BryceMason.

Examples:
http://www.mu-sigma.com/
http://www.salford-systems.com/
http://www.angoss.com/

 
  • IP logged

g
  • Posts: 15
Regression Trees
#1: June 05, 2014, 11:00:00 PM
I have always thought that a neural net was the best way to build loan prediction systems.

http://www.r-bloggers.com/using-neural-networks-for-credit-scoring-a-simple-example/
  • IP logged

  • Posts: 300
Regression Trees
#2: June 05, 2014, 11:00:00 PM
PeerCube uses Decision Tree/Regression Tree for splitting the loan attributes before calculating BLE Risk Index. If you are familiar with R, you can use it with 'party' or 'rpart' packages. As the example below and attached shows, even with three loan attributes and ROI, the analysis gets complex.

Personally, I suggest lenders look into Genetic Algorithm as described by David M. Patierno http://blog.dmpatierno.com/post/3161338411/lending-club-genetic-algorithm. PeerCube has it as Public Loan Filter 'DMP Genetic Algorithm'.

Code: [Select]
  • IP logged

b
  • Posts: 152
Regression Trees
#3: June 05, 2014, 11:00:00 PM
It can work well, but I'm not a fan. I prefer to understand the reasons why we see relationships, and the computer finding the best way to slide down regression trees / random forests like a game of Plinko on the Price is Right just doesn't satisfy my curiosity. There are many ways to skin a cat.
  • IP logged

T
  • Posts: 6588
Regression Trees
#4: June 05, 2014, 11:00:00 PM
  • IP logged

F
  • Posts: 282
Regression Trees
#5: June 07, 2014, 11:00:00 PM
  • IP logged

T
  • Posts: 6588
Regression Trees
#6: June 07, 2014, 11:00:00 PM
That's a fine approach to improve models, but one better use a very high alpha because in my book this is a fishing expedition. You check every permutation with a genetic or tree and there are bound to be (1-alpha)% variables significant just by chance. So, back testing and high alpha are important here IMO.
  • IP logged

T
  • Posts: 6588
Regression Trees
#7: June 07, 2014, 11:00:00 PM
  • IP logged

Regression Trees
#8: June 09, 2014, 11:00:00 PM
  • IP logged


n
  • Posts: 1
Regression Trees
#10: November 12, 2019, 12:00:00 AM
even i think that a neural net was the best way to build loan prediction systems

example: https://www.nuncsystems.com/big-data.html
  • IP logged