Question
Types of predictive models available in PAD
As far as I know, there are currently two main types of predictive models available in PAD:
- scoring (binary output)
- spectrum (prediction of continuous behavior)
However, would it be possible to create one single model for dependent variable having multiple classes, i.e.: predicting propensities of purchase for multiple products, or multiple models would need to be built for each product in order to get individual score for single products?
Thank you.
Thank you for your response.
Does it mean I can predict probabilities of multiple categories (in my example: products to be purchased) using ONE SINGLE MODEL i.e. propensity of purchasing product 1, product 2 etc.?
Which type of model would it be cause scoring needs to have binary outcome (product 1/not product 1) and spectrum works with continuous variable only..?
Can you please clarify it?
Thank you
Do you have people on your team with data science background? Here is the Pega Academy course you might want to explore: https://academy.pega.com/library/8/pega-data-scientist. Feature research/development is not unique to Pega.
Hi,
Yes. Scoring model provides binary outcome. Spectrum model is for continuous outcome.
Thank you.
Thank you all.
I am aware what scoring and spectrum models stand for. Just wanted to get confirmation that there is no other with multi class dependant variable available. From what I understand it is not.
Btw, I am a data scientist and I already have been using PAD;)
Hi,
You are completely correct, as PAD is for binary scoring it can only cover a binary outcome and as a scoring model it produces a probability for being positive. For multiclass scoring in PAD you would create one model per product (product x versus the rest).
As an aside for product and propensity modeling typically online learning, ie 'Adaptive Decisioning' is being used. It will learn on the fly for every outcome being captured, and creates one model per proposition. When the system encounters any new propositions it will create a new model on the fly, and models are constantly learning.
Earlier in the trail extended scoring models were being mentioned, but these do not relate to multiclass classification. An extended scoring model is a very special model that you use if for a non select part of your population you dont have the outcome. The prototypical use case is credit scoring, where you dont have behavior (default y/n for customers that were declined or that declined your offer). This is also called reject inference or outcome inferencing.
You are right, those are two main types of predictive models in Pega (there is another called extended scoring model which requires separate license (reference: https://community.pega.com/sites/default/files/help_v81/pxpredictionstudio/procomhelpmain.htm#dsm/da-portal/concepts/da-predictive-models-con.htm). I don't see why you cannot use 'product' as one of the dependent variables (or feature in data science term) for either model types.