This repository offers a repertoire of rudimentary neural networks and deeplearning implementations developed with Python. It is intended for people who seek to study the basics of neural networks and deeplearning.
Planar Data Classification with One Hidden Layer This project demonstrates basic neural network and deep learning implementations in Python. It covers defining the neural network structure, initializing parameters, forward propagation, cost computation, backpropagation, parameter updates, and integrating these to build and test models. Additionally, it enables hidden layer tuning to analyze different behaviors and uses a planner database for evaluation. This serves as an excellent starting point for understanding fundamental neural network concepts.
Developing a Deep Neural Network: A Comprehensive Guide This project starts with establishing network architecture, setting up parameters, forward propagation, calculating expenses, backpropagation, updating parameters, creating the neural network model, and testing predictions are all illustrated in the projects. It also looks at adjusting hidden layers to see different behaviours.
Deep Neural Network - Application- Cat Image Classification This project uses a deep neural network to classify "cat" photos. It builds upon prior projects to construct and train both a two-layer neural network and an L-layer deep neural network. A systematic approach—initializing parameters, forward propagation, cost computation, backpropagation, and parameter updates—improves logistic regression accuracy. The project also tests the model with custom photographs and analyses outcomes to discover issues like unexpected image angles or backdrops. It provides a complete introduction to supervised learning neural network construction and training.