MatLogica AADC

Summary:
MatLogica AADC approach uses Code Transformation and Operator Overloading to efficiently compute the gradients of mathematical models by generating optimized machine code at runtime. This results in faster computing of the model and its first and higher order derivatives. The approach is particularly useful for models with many parameters that require frequent gradient updates during training.

URL: https://matlogica.com/index.php

Developers:

• Dmitri Goloubentsev
• Evgeny Lakshtanov

Mode:	Forward Reverse
Method:	Source transformation Operator overloading
Supported Language:	C# C/C++ Python

Awards

4 'Category Leader' awards from Chartis in 2023: Innovation; AAD; Data Parallel
Programming; Innovation in Computational Frameworks and ranked #10 out of 50
in the QuantTech rating.

Supported Platforms:

• Windows
• Unix/Linux

Licensing: free with restrictions

Entries in our publication database that actually use MatLogica AADC in the numerical experiments:  0

The following diagram shows these entries versus the year of the publication.

10+ #Entries 0
	Year