CS-702 (B) Deep & Reinforcement Learning Notes
5 notes for Deep & Reinforcement Learning, unit by unit.
Notes for CS-702 (B)
- Unit 1 · Short notes Foundations of Deep Learning History of Deep Learning · McCulloch‑Pitts neuron · Thresholding Logic · Activation functions · Gradient Descent (GD) · Momentum based GD
- Unit 2 · Short notes Autoencoders and Regularization Autoencoders and their relation to PCA · Regularization in autoencoders · Denoising autoencoders · Sparse autoencoders · Contractive autoencoders · Bias–Variance trade‑off
- Unit 3 · Short notes Convolutional Architectures and Pre‑training Greedy layerwise pre‑training · Advanced activation functions · Weight initialization methods · Learning vector representations of words · Convolutional Neural Networks (CNNs) · LeNet
- Unit 4 · Short notes Fundamentals of Reinforcement Learning Introduction to Reinforcement Learning (RL) · Bandit algorithms – UCB, PAC, Median Elimination · Policy Gradient methods · Full RL & Markov Decision Processes (MDPs) · Bellman optimality principle · Dynamic programming – value iteration, policy iteration
- Unit 5 · Short notes Advanced RL Techniques Fitted Q · Deep Q‑Learning (DQN) · Advanced Q‑learning algorithms · Learning policies by imitating optimal controllers · Policy Gradient algorithms for full RL · Hierarchical RL
More for CS-702 (B)
Other 7th semester CSE subjects
- CS-701 Software Architectures
- CS-702 (A) Computational Intelligence
- CS-702 (C) Wireless & Mobile Computing
- CS-702 (D) Big Data
- CS-703 (A) Cryptography & Information Security
- CS-703 (B) Data Mining and Warehousing
- CS-703 (C) Agile Software Development
- CS-703 (D) Disaster Management