Subject Syllabus
CS-702 (B) - Deep & Reinforcement Learning
CSE · Semester 7
Course Modules
Learn topic-wise with Hyper AI and revise using mapped PYQs.
History of Deep LearningMcCulloch Pitts Neuron, Thresholding Logic, Activation functionsGradient Descent (GD), Momentum Based GD, Nesterov Accelerated GD, Stochastic GD, AdaGrad, RMSProp, AdamEigenvalue DecompositionRecurrent Neural Networks, Backpropagation through time (BPTT)Vanishing and Exploding Gradients, Truncated BPTTGRU, LSTMsEncoder Decoder Models, Attention Mechanism, Attention over images
Autoencoders and relation to PCARegularization in autoencoders, Denoising autoencoders, Sparse autoencoders, Contractive autoencodersRegularization: Bias Variance Tradeoff, L2 regularization, Early stoppingDataset augmentation, Parameter sharing and tying, Injecting noise at inputEnsemble methods, DropoutBatch Normalization, Instance Normalization, Group Normalization
Greedy Layerwise Pre-trainingBetter activation functions, Better weight initialization methodsLearning Vectorial Representations Of WordsConvolutional Neural NetworksLeNet, AlexNet, ZF-Net, VGGNet, GoogLeNet, ResNetVisualizing Convolutional Neural Networks, Guided BackpropagationDeep Dream, Deep ArtRecent Trends in Deep Learning Architectures
Introduction to reinforcement learning (RL)Bandit algorithms - UCB, PAC, Median EliminationPolicy GradientFull RL & MDPs, Bellman OptimalityDynamic Programming - Value iteration, Policy iterationQ-learning & Temporal Difference MethodsTemporal-Difference Learning, Eligibility TracesFunction Approximation, Least Squares Methods
Fitted Q, Deep Q-Learning, Advanced Q-learning algorithmsLearning policies by imitating optimal controllersDQN & Policy Gradient, Policy Gradient Algorithms for Full RLHierarchical RL, POMDPs, Actor-Critic MethodInverse reinforcement learningMaximum Entropy Deep Inverse Reinforcement LearningGenerative Adversarial Imitation LearningRecent Trends in RL Architectures