What are Parsimony Nontriviality in credit scoring models?

[frmpart2dan]

Could someone explain what is Parsimony and Nontriviality requirement in credit scoring model means?

 

[QuantMan2318]

I would personally think that any credit scoring model is basically the same as the models that are used in Machine Learning, the word parsimony, as a scientific principle refers to something that can be done in the most simplest of ways, therefore, in the same vein, we refer to models in Machine Learning as well as pricing to have parsimony when they can compute the prices or explain the trend in the data with the minimum number of independent variables or parameters (Nefti states the same in his beautiful book on Financial Engineering)

I would presume that credit scoring models have to predict the worthiness of the credit with the minimum number of parameters that can explain most of the situations or else the models can become extremely complex

I don't exactly know what non trivial is, but interpolating based on what Servigny says, we can assume that the model should produce results with a reasonable degree of confidence or reliability and should produce results that are difficult to prove and not those that can be verified easily

 

[brian.field]

I agree with @QuantMan2318 (all would do well to do so!)

I think of parsimonious as Occam's Razor and I think of triviality as the Interocular Trauma Test!

 

[David Harper CFA FRM]

These are among the five requisite qualities cited by de Servigny, I have copied them below. I retrieved the referenced paper (Galindo and Tamayo, here it is at https://www.dropbox.com/s/5kiwrb5seji2rs7/galindo-tamayo-credit-scoring.pdf?dl=0); however, I actually do not see these criteria articulated in the paper, sorry Maybe I did not scan the paper well (?)

"Choosing the optimal model, based on an existing data set, remains a real challenge today. Galindo and Tamayo (2000) have defined five requisite qualities for the choice of an optimal scoring model:

1. Accuracy. Having low error rates arising from the assumptions in the model
   
2. Parsimony. Not using too large a number of explanatory variables
   
3. Nontriviality. Producing interesting results
   
4. Feasibility. Running in a reasonable amount of time and using realistic resources
   
5. Transparency and interpretability. Providing high-level insight into the data relationships and trends and understanding where the output of the model comes from -- "Servigny, Arnaud de; Olivier Renault. Measuring and Managing Credit Risk (Standard & Poor's Press) (Kindle Locations 1298-1305). McGraw-Hill Education. Kindle Edition."