Search this site

Deep Analytics

Brian Huge & Antoine Savine

Superfly Analytics, Danske Bank

DeepAnalyticsNov19short

Deep Learning Black & Scholes, from Deep Analytics by Brian Huge & Antoine Savine

Neural Net learning Black & Scholes from samples

Deep Learning Basket Options, from Deep Analytics by Brian Huge & Antoine Savine

Neural Net learning Basket Options from samples

The presentation heavily relies on differentiation, and AAD, an algorithm to compute vast numbers of differentials very quickly and exactly, to machine precision. Readers without knowledge of AAD may still read the presentation, think of AAD as a black-box algorithm to compute differentials quickly. This being said, we regard AAD as the single most significant addition to numerical finance since the introduction of finite difference methods (FDM) in the early 1990s. We encourage students, academics and practitioners to learn this formidable technology, which recently changed the face of derivatives finance. You may find information and learning resources for AAD and application to finance here:

https://sites.google.com/view/antoinesavine-aad/

Or watch the 15min video below, introducing AAD and backpropagation in Machine Learning and Finance:

Slides here:

aad15min.pdf

Winner: In House System of the Year 2015 Risk award

Winner: Excellence in Risk Management and Modelling 2019 RiskMinds award

See overview of the platform below

Riskminds_28112919_2.pdf

Brian Huge

brian.huge@deep-analytics.org

Antoine Savine

antoine.savine@deep-analytics.org

Report abuse

Page details

Page updated

Report abuse