A bio-plausible alternative to the standard Backpropagation algorithm for deep learning. It implements Difference Target Propagation to train neural networks without traditional gradient descent.
Tech Stack
PythonAI, ML & Data
Why COMP-490-Research-Project?
• Bio-plausible architecture
• Comparative evaluation metrics
• Clean NumPy implementation
Limitations
• Lower accuracy than BP
• Experimental status
• Limited layer support
7/1/2025
Last Update
0
Forks
0
Issues
MIT
License
Financial Leak Detected
Stop the "SaaS Tax"
Your team could be burning cash. Switching to COMP-490-Research-Project instantly boosts your runway.
