Learn How to Enhance Your Interest Rate Models with Machine Learning
Download the Paper
Download Alexander Sokol’s working paper “Autoencoder Market Models for Interest Rates” and explore how classical interest rate models in the Q- and P-measures can be modified to use variational autoencoders (VAE) for dimension reduction.
What We Do
ModVal.org is an online resource for model validation.
ModVal.org is free of charge, but requires online
registration.
View source and data files
Download raw data files
Getting Started
Register to use Modval.org and get access to
annotated tests
in our extensive collection of financial models and calculations.
Model Validation
Utilize Modval.org’s source code to work on validating your models, use your inputs
to
visualize data, create plots, and trace the impact of calculation method choices.
Contribute to ModVal.org
ModVal.org releases source code for the purpose of fostering
collaboration between trading desk and risk management practitioners,
industry experts, and academic researchers working in the field of Quantitative
Finance and Risk Management.
Published Research
Reproduce and review every step of the calculations from the latest research papers,
including
“Rethinking Margin Period of Risk” by Leif
Andersen, Michael Pykhtin, and Alexander Sokol.
