The Neural Networks course introduces second-year BSc Artificial Intelligence students to the fundamental principles and architectures underlying modern neural computation. The course begins with foundational concepts such as perceptrons, multi-layer perceptrons, and the backpropagation algorithm. It then develops the student’s understanding of deep learning architectures including convolutional neural networks, recurrent neural networks, and transformers. In the final weeks, the course explores biologically inspired models such as Hebbian learning, Hopfield networks, and Boltzmann Machines. Students will gain both theoretical insight and practical experience through lectures, hands-on sessions, and supervised training exercises, preparing them for more advanced AI coursework and research.
This course introduces the foundational principles and architectures of neural networks, from simple perceptrons to advanced deep learning systems. Students begin by studying historical developments and core concepts such as activation functions, layers, and the backpropagation algorithm. The course covers optimization and regularization techniques essential for training neural networks effectively. Subsequently, it introduces convolutional neural networks (CNNs), residual networks, recurrent neural networks (RNNs), and transformer architectures. In the final weeks, biologically inspired models such as the Hebbian learning rule, Hopfield networks, and Boltzmann machines are presented. The course balances theoretical understanding with good practices in training and evaluating networks, preparing students for both academic research and real-world applications in AI.