Deep Learning and Reinforcement Learning

This module introduces Deep Learning and Neural Networks fundamentals, covering neuron basics, forward propagation, matrix representation, and different types of neural networks.

This module covers the mathematics of Back Propagation, activation functions, and practical implementation using Keras. Students learn about different activation functions and their roles in introducing non-linearity to networks.

This module focuses on optimization techniques for neural networks, including different optimizers, momentum, regularization, and data shuffling. It provides hands-on experience with Keras and grid search techniques.

This module introduces CNNs, covering image processing concepts like kernels, convolution operations, padding, stride, and pooling. It includes practical implementations and hands-on exercises with CNN architectures.

This module covers transfer learning concepts and implementation using pre-trained CNN architectures like VGG-16 and ResNet-50. It includes detailed study of various CNN architectures and regularization techniques.

This module focuses on RNNs and LSTM networks, covering their architecture, applications in speech-to-text recognition, and implementation details. It includes practical exercises with sequential data processing.

This module covers autoencoder architecture and applications in unsupervised learning, focusing on dimensionality reduction and image applications. It includes practical implementation exercises.

This module explores generative models including VAEs and GANs, their implementation in Keras, and GPU-accelerated training. It covers practical aspects of generating artificial images.

This final module introduces reinforcement learning concepts and applications, focusing on reward-based training approaches instead of traditional error minimization methods. It includes a final project and peer review component.