Pyrenn LM Neural Network Training Algorithm Bench Test

Version 1.0.0 (21.4 KB) by Mike Fouche
The code allows you to test the Pyrenn LM algorithm against Matlab's LM algorithm for the quadratic function.
76 Downloads
Updated 2 Feb 2021

View License

The original Pyrenn code base (created by Dennis Atabay and his colleagues) - available as Matlab scripts or Python, can be downloaded here - https://pyrenn.readthedocs.io/en/latest/.

I used their code (and communicated with Dennis), to develop a simple framework to compare performance of their Neural Network Levenberg-Marquardt (LM) learning algorithm against Matlab's LM learning algorithm for a Feed-Forward Neural Network architecture.

The Pyrenn algorithm works fast and performs very, very well. My write-up on the comparison can be read here - https://mikescodeprojects.com/2020/01/12/pyrenn-vs-matlab/.

You must have the Matlab Deep Learning toolbox in order to run the Matlab LM algorithm for comparing performance. But if you don't, you can still run the Pyrenn algorithm (the code base is included in the package for this post) and adapt it for your requirements (other applications).

Here is a video of how to run the package in Matlab - https://mikescodeprojects.files.wordpress.com/2020/01/running_in_matlab.mp4?_=1.

This is a video of a code walk-through - https://mikescodeprojects.files.wordpress.com/2020/01/code_walk-through.mp4?_=3.

Cite As

Mike Fouche (2024). Pyrenn LM Neural Network Training Algorithm Bench Test (https://www.mathworks.com/matlabcentral/fileexchange/86790-pyrenn-lm-neural-network-training-algorithm-bench-test), MATLAB Central File Exchange. Retrieved .

MATLAB Release Compatibility
Created with R2017b
Compatible with any release
Platform Compatibility
Windows macOS Linux

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!
Version Published Release Notes
1.0.0