For a specific job type, I have a .csv table with applicant data storing whether or not the applicant has necessary skills for the role as binary variables. The outcome, whether or not the applicant received an interview, is also stored as a binary classifier. I was hoping to create an algorithm using a Predictive Model Decision Shape that takes in information collected from a Collect Information step that maps to the features of the .csv about an applicant, and outputs an interview probability between 0 and 1.
Is there a way to produce a continuous variable probability output based on an algorithm that uses a table whose outcomes are binary classifiers with a Predictive Model Decision Shape, as described above? If so, how would you do this?
In the strategy first, call a sub-strategy place an adaptive model where it will classify the binary outputs and outputs are written to a db table once that is done in the main strategy you embed the data from the table you saved the results of the sub-strategy and put another adaptive model using the regression to estimate the continuous variable information.
But make sure that the model in the adaptive model is tunned first correctly.
Posted: 4 years ago
Posted: 26 Jul 2019 4:30 EDT
Otto Perdeck (Otto_Perdeck)
Director, Data Science, Machine Learning & AI
Sounds like a classical use case to create a predictive model using Pega Machine Learning - in previous product versions "PAD". You have a historical data set so use of Adaptive is less of a natural fit although you could stream the data in to Adaptive as well.
Create the model in Prediction Studio, then embed it in your strategy to get a real-time score that you can act upon.
If you're unfamiliar with these please check out the academy material on Decisioning / AI.