Deep Arbitrage Free Regularization

This code is accompaniment for the paper "Deep Learning in a Generalized HJM-type Framework through Arbitrage-Free Regularization" by A. Kratsios and C. Hyndman; available on ArXiV.

How to Use:

First:

Read/Import data-set to bond.data

Second

Run Main_NS or Run Main_dPCA depending on which model to test.

Important:

Both the Nelson-Siegel and dynamic PCA experiments should be run separately.