Warsaw University of Technology Neural Network Course Practice: Six Mini-Projects to Help You Deeply Understand the Core of AI

The open-source repository of six mini-projects from Warsaw University of Technology's SSNE (Artificial Neural Networks) course, maintained by hbiegacz, covers a complete learning path from basic perceptrons to convolutional neural networks, recurrent neural networks, and optimization techniques, providing a systematic hands-on guide for deep learning beginners.

Warsaw University of Technology is a top technical university in Poland, and its SSNE course adopts a dual-track model of "theory + practice". The learning objectives are not only to master basic concepts of neural networks but also to focus on cultivating practical skills. Each project targets a specific topic, guiding students to deeply understand the core mechanisms of AI.

The repository contains six projects:

1. Neural Network Basics and Perceptrons: Implement a binary classifier, understand weight updates and gradient descent;
2. Multi-layer Feedforward Networks and Backpropagation: Implement the backpropagation algorithm by hand;
3. Introduction to Convolutional Neural Networks: Learn the principles of convolutional layers and pooling layers;
4. Recurrent Neural Networks and Sequence Modeling: Explore RNNs and their variants for processing time-series data;
5. Regularization and Optimization Techniques: Implement Dropout, batch normalization, and various optimizers;
6. Experimental Materials and Supplementary Resources: Provide datasets, reference papers, and other auxiliary materials.

The course project design embodies best practices:

• Progressive difficulty curve: Avoid knowledge gaps;
• Hands-on first: Each concept is accompanied by runnable code;
• Covers core paradigms: Feedforward, convolutional, recurrent networks, and training optimization techniques.

The open-source repository provides a structured self-learning path for enthusiasts who cannot enroll in the course. The code uses Python and mainstream frameworks (PyTorch/TensorFlow). After completing the projects, learners will have the ability to independently design and train neural network models, preparing them for complex AI projects or research.

This set of course projects demonstrates the core elements of high-quality AI education: clear structure, step-by-step progression, and sufficient practice. It is suitable for computer science students, career-changers, and AI self-learners. Following the projects allows learners to find their learning rhythm, and true understanding comes from hands-on building.