Neural Networks
| Course Code | COMP-258 |
|---|---|
| Lecture hours per week | 2 |
| Lab hours per week | 2 |
| Course Availability | Open |
| Description | This course explores artificial neural networks and their applications in solving real-world problems. Students will study foundational models such as McCulloch-Pitts neurons, the Hebbian rule, Perceptrons, and Multi-Layer Perceptrons (MLPs), as well as advanced architectures including Convolutional Neural Networks (CNNs), Vision Transformers (ViTs), state-of-the-art in large-scale image recognition, and models for sequence processing, such as Recurrent Neural Networks (RNNs), Long Short-Term Memory networks (LSTMs), and Transformers. Key topics include backpropagation, activation functions, regularization, optimization techniques, embedding layers, sequence modelling, and attention mechanisms. Students will gain hands-on experience with Keras and TensorFlow to design, train, and evaluate neural networks, and will apply them to tasks in image recognition, natural language processing (NLP), and generative AI. AI-assisted development tools are introduced to enhance coding efficiency and model performance. |