The course covers the area of neural networks with reference to other techniques from the broader area of computational intelligence.
It explores neural network models and architectures, dynamic behavior, convergence and stability, learning algorithms, implementations, computational capabilities, and applications. Feed-forward networks and learning through error correction (multi-layer perceptron and backpropagation). Support vector machines (SVM). Associative networks, Hopfield networks, recurrent multilayer networks. Competitive learning and Kohonen maps. Combinatorial optimization algorithms. Genetic algorithms. Deep learning: convolutional neural networks (CNNs), recurrent neural networks (RNNs), autoencoders, generative adversarial networks (GANs). Reinforcement learning: dynamic programming, value iteration, Q-learning, deep Q-learning. Fuzzy logic and knowledge engineering. It also comprises laboratories on supervised learning, unsupervised learning, deep learning, and reinforcement learning.