This repository contains implementations of deep neural networks (DNNs) with 1 layer, 3 layers, and a generalized L-layer model, inspired by Coursera-style deep learning assignments.

Files

• one_layer_dnn.ipynb: A shallow neural network with one hidden layer.
• three_layer_dnn.ipynb: A 3-layer neural network with two hidden layers.
• l_layer_dnn.ipynb: A generalized L-layer neural network for flexible depth.
• cat_image_dnn.ipynb: A generalized L-layer neural network for cat image classification.

Requirements

• Python 3.x
• NumPy
• Matplotlib

How to Use

1. Clone this repository: git clone <your-repo-url>
2. Open the Jupyter notebooks in VS Code or Jupyter Notebook.
3. Run the cells to train the models on synthetic binary classification data.
4. Observe the cost reduction plots.

Notes

• The data is synthetic (2 features, 400 samples) for binary classification.
• Adjust hyperparameters (learning rate, epochs, layer sizes) as needed.
• Models use ReLU for hidden layers and sigmoid for output.