This repository is part of Deep Learning 1, the first course in a two-part deep learning specialization. It contains four exercises completed during the first semester of the Artificial Intelligence Master’s program at University of Amsterdam.
Implements an image classifier using only NumPy, without relying on standard deep learning libraries like PyTorch.
This approach involves manually constructing a neural network architecture and mathematical operations, explicitly showcasing underlying linear algebra operations.
Tackles the same image classification task using the PyTorch library.
This project demonstrates how to leverage PyTorch's high-level functions to efficiently construct, train, and evaluate neural network models.
By utilizing PyTorch, this project provides a contrast to the manual implementations of the NumPy-based classifier.