Systems and methods for predictive modeling
Discussed herein are methods and systems for an interdependent series/suite of AI models. In one embodiment, a processor receives projection assumption inputs from a user device and executes a population builder machine learning model to predict a dynamic adjustment table. It applies the model to population data to generate a value population file, which simulates a subset of the population based on the predicted table. The file contains value cells representing instances of a product. The device then runs a mortality machine learning model to determine mortality data for the product using the simulated population. Finally, it executes a flow projection model to generate a projection report for the product, incorporating mortality data and projection assumptions.
1. A system for using a suite of interdependent artificial intelligence models to improve each individual model's efficiency, the system comprising:
a non-transitory computer-readable medium storing a computer program instructions, and a processor coupled to the non-transitory computer-readable medium and configured to execute the instructions to:
in response to receiving a plurality of projection assumption inputs from a user device,
execute a population builder machine learning model to:
predict a dynamic adjustment table in accordance with the assumption inputs, and
output a value population file representing a simulated population by applying the population builder model to population data, wherein the value population file comprises a set of value cells representing instances of a product, the simulated population corresponding to a subset of a population selected in accordance with the dynamic adjustment table predicted by the population builder machine learning model;
execute a mortality machine learning model to determine mortality data for the product by ingesting the value population file generated by the population builder machine learning model;
execute a flow projection model to output for display at the user device a projection report for the product by applying the flow projection model to the mortality data ingested from the mortality machine learning model in addition to projection assumptions data based on the plurality of projection assumption inputs, wherein the projection assumptions data represent assumptions for the flow projection model; and
generate a product file comprising code that represents the product and the projection report.
2. The system of claim 1 , wherein the processor is further configured to execute the instructions to:
generate the population data in response to receiving a plurality of population inputs from the user device, wherein the population data represent value cell specifications associated with the product.
3. The system of claim 2 , wherein the population data comprise a plurality of cell keys, wherein respective cell keys index one or more of the value cell specifications to cell identities in the population data.
4. The system of claim 1 , wherein the projection assumptions data comprises valuation parameters associated with the flow projection model.
5. The system of claim 4 , wherein the valuation parameters include decrement parameters representative of mortality and lapse decrements.
6. The system of claim 1 , wherein the projection assumptions data comprise product features and market scenarios.
7. The system of claim 1 , wherein the set of value cells represent inforce instances of the product, and the set of value cells include weightings of respective pricing cells in the simulated population.
8. The system of claim 1 , wherein the mortality data for the product comprise adjusted mortality variables including population based mortality adjustments based on the simulated value population.
9. The system of claim 8 , wherein the adjusted mortality variables comprise adjusted mortality tables including population based table adjustment surfaces determined via analysis of the value population file.
10. The system of claim 1 , wherein the flow projection model is communicatively coupled with the machine learning model via a call-response interface that handles API (application programming interface) calls from the flow projection model and responses from the machine learning model.
11. The system of claim 1 , wherein the processor is further configured to execute the instructions to:
train the machine learning model by applying survival modeling to a plurality of historical application records.
12. The system of claim 1 , wherein the processor is further configured to execute the instructions to:
execute the machine learning model to generate value prediction data representative of one or both of a placement rate and a premium persistency rate; and
execute the flow projection model to output the projection report by applying the flow projection model to the value prediction data.
13. The system of claim 1 , wherein the flow projection model is further configured to execute a profitability solver, wherein the projection report comprises profitability data.
14. A method for using a suite of interdependent artificial intelligence models to improve each individual model's efficiency, the method comprising:
receiving, by at least one processor, a plurality of projection assumption inputs from a user device;
executing, by the at least one processor, a population builder machine learning model to:
predict a dynamic adjustment table in accordance with the assumption inputs, and
output a value population file representing a simulated population by applying the population builder model to population data, wherein the value population file comprises a set of value cells representing instances of a product, the simulated population corresponding to a subset of a population selected in accordance with the dynamic adjustment table predicted by the population builder machine learning model;
executing, by the at least one processor, a mortality machine learning model to determine mortality data for the product by ingesting the value population file generated by the population builder machine learning model; and
executing, by the at least one processor, a flow projection model to output for display at the user device a projection report for the product by applying the flow projection model to the mortality data ingested from the mortality machine learning model in addition to projection assumptions data based on the plurality of projection assumption inputs, wherein the projection assumptions data represent assumptions for the flow projection model; and
generating, by the at least one processor, a product file comprising code that represents the product and the projection report.
15. The method of claim 14 , further comprising:
generating, by the at least one processor, the population data in response to receiving a plurality of population inputs from the user device, wherein the population data represent value cell specifications associated with the product.
16. The method of claim 15 , wherein the population data comprise a plurality of cell keys, wherein respective cell keys index one or more of the value cell specifications to cell identities in the population data.
17. The method of claim 14 , wherein the projection assumptions data comprises valuation parameters associated with the flow projection model.
18. The method of claim 17 , wherein the valuation parameters include decrement parameters representative of mortality and lapse decrements.
19. The method of claim 14 , wherein the projection assumptions data comprise product features and market scenarios.
20. The method of claim 14 , wherein the set of value cells represent inforce instances of the product, and the set of value cells include weightings of respective pricing cells in the simulated population.