This is a basic deep neural network developed from scratch for a better grasp of neural network mechanics. It's applied to image classification for cat and non-cat images.
The L-layer Model currently yields low accuracy at 34%. This may stem from issues in the linear_activation_backward() function or suboptimal hyperparameters. I'm working on resolving this. My apologies.
In contrast, the 2-layer Model achieves a higher accuracy of around 72%.