CS 4/5789: Introduction to Reinforcement Learning

Lecture 11

Prof. Sarah Dean

MW 2:45-4pm
110 Hollister Hall



0. Announcements & Recap

1. Supervision via Rollouts

2. Approximate Policy Iteration

3. Performance Different Lemma

4. Conservative Policy Iteration


HW0 grades and solutions released

HW1 due Monday 3/7

5789 Paper Review Assignment (weekly pace suggested)

Prelim Tuesday 3/22 at 7:30pm in Phillips 101

Office hours after lecture M (110 Hollister) and W (416A Gates)

Wednesday OH prioritize 1-1 questions/concerns over HW


Model-based RL (Meta-Algorithm)

  1. Sample and record \(s_i'\sim P(s_i, a_i)\)
  2. Estimate \(\widehat P\) from \(\{(s_i',s_i, a_i)\}_{i=1}^N\)
  3. Design \(\widehat \pi\) from \(\widehat P\)

Tabular MBRL

  1. Sample: evenly each \(s,a\) \(\frac{N}{SA}\) times
  2. Estimate: by averaging
  3. Design: policy iteration

Using Simulation Lemma we saw that \(\epsilon\) optimal policy requires \(N\gtrsim \frac{S^2 A}{\epsilon^2}\)


  1. Sample: \(s_i\sim \mathcal N(0,\sigma^2 I)\) and \(a_i\sim \mathcal N(0,\sigma^2I)\)
  2. Estimate: by least-squares linear regression
    \(\displaystyle \widehat A,\widehat B = \arg\min_{A,B} \sum_{i=1}^N\|s_i' - As_i-Ba_i\|_2^2\)
  3. Design: LQR dynamic programming


We will not cover the argument in detail, but \(\epsilon\) optimal policy requires \(N\gtrsim \frac{n_s+n_a}{\epsilon^2}\)

Recap: Policy Iteration

Policy Evaluation uses knowledge of transition function \(P\) and reward function \(r\)

initialize pi[0]
for t=1,2,...
    Q[t] = PolicyEvaluation(pi[t]) # Evaluation
    pi[t+1] = argmax_a(Q[t](:,a)) # Improvement

Can we estimate the Q function from sampled data?

CS 4/5789: Lecture 11

By Sarah Dean


CS 4/5789: Lecture 11