RTMNU · Final Year B.Tech Information Technology · CBCS
Reinforcement Learning
BTIT801T.2 · Elective VI · Semester VIII · 3 credits · College 30 + university 70
One of the Elective VI options. An NPTEL course: its page lists twelve weeks, grouped here into five units.
Unit 1: Introduction and Bandit Algorithms (weeks 1–3)
What reinforcement learning is; bandit algorithms — UCB, PAC bounds and median elimination — and policy-gradient methods for bandits.
Examined in the 70-mark end-semester paper
- Elements of reinforcement learning
- The n-armed bandit problem
- Optimistic initial values and UCB action selection
- Gradient bandits
- PAC bandit algorithms and median elimination
Unit 2: Full Reinforcement Learning and MDPs (weeks 4–5)
The full RL problem as a Markov decision process; Bellman optimality.
Examined in the 70-mark end-semester paper
- The agent–environment interface
- Markov decision processes
- Value functions and optimal value functions
Unit 3: Dynamic Programming, TD Methods and Eligibility Traces (weeks 6–7)
Dynamic programming for RL; temporal-difference methods; eligibility traces.
Examined in the 70-mark end-semester paper
- Policy iteration
- Value iteration
- Temporal-difference prediction
- Q-learning and SARSA
- Eligibility traces and TD(λ)
Unit 4: Function Approximation and Deep RL (weeks 8–10)
Function approximation and least-squares methods; fitted Q-iteration, deep Q-networks and policy gradients for full RL.
Examined in the 70-mark end-semester paper
- Value-function approximation
- Semi-gradient methods and linear function approximation
- Deep Q-networks
- Policy gradient theorem
Unit 5: Hierarchical RL and POMDPs (weeks 11–12)
Hierarchical reinforcement learning; partially observable Markov decision processes.
Examined in the 70-mark end-semester paper
- Hierarchical reinforcement learning
- Partially observable MDPs
Unit contents are summarised from the official RTMNU syllabus . AiCanCode.org is not affiliated with Rashtrasant Tukadoji Maharaj Nagpur University (RTMNU).