How to find a right machine learning algorithm for the given problem?


#1

How To find the right Machine learning algo? I have one problem statement as follows:

Based on the Defect summary, how would I recommend the set of regression test case set?

Can anyone show me the path to get started with the problem statement?


#2

Simplest way is use the data miner application in R. You get some models to start with. There is no right machine learning alo. There is always a right way to interpret the output of an algo


#3

Hello,

When you have to model “counts” (a Defect report will give you details about how many defects by category you get) the way to go is to use a GLM (Generalized Linear Model) where your link function (R’s terminology) is “poisson”.

fit <- glm(defects ~ your_predictors, data = your_data_frame , link = “poisson”).

Regards,
Carlos Ortega - Spain