Can an insurance pricing model improve predictive accuracy without becoming a black box?
Interpretable Multi-Task Insurance Pricing
Built an interpretable PyTorch multi-task neural additive model using shared feature subnetworks to jointly predict claim frequency and severity on freMTPL data. Evaluated the model with six-fold cross-validation against GLM, GBM, and single-task NAM benchmarks.
PythonRPyTorchMulti-Task LearningNeural Additive Models
View Project 
