- Handbook of Learning and Approximate Dynamic Programming - Google Books
- Publications
- Navigation Bar
- Handbook of Learning and Approximate Dynamic Programming

How to Manage Your Award. Grant Policy Manual. Grant General Conditions.

## Handbook of Learning and Approximate Dynamic Programming - Google Books

Cooperative Agreement Conditions. Special Conditions. Federal Demonstration Partnership. Policy Office Website. NSF Org:. Jennie Si si asu. Program Reference Code s :.

- Handbook of Learning and Approximate Dynamic Programming - tomiddperspafi.gq?
- Knowledge Processing with Interval and Soft Computing.
- New Approaches in Intelligent Control: Techniques, Methodologies and Applications.
- A Quick Introduction to ADPRL;
- Warren B Powell - Google 學術搜尋引用文獻.
- Pok Pok: Food and Stories from the Streets, Homes, and Roadside Restaurants of Thailand.

Some of these topics are large, so students can choose some suitable subset on which to lead a discussion. Topics that we will definitely cover eg: I will lead the discussion if nobody else wants to : Q-learning and Temporal-Difference Learning. Rollout, limited lookahead and model predictive control. Optimal control in continuous time and space.

LQR for linear optimal control. Kalman filters for linear state estimation. The Pontryagin minimum principle. Eikonal equation for shortest path in continuous state space and the Fast Marching Method for solving it. Topics that we will cover if somebody volunteers eg: I already know of suitable reading material : DP for financial portfolio selection and optimal stopping for pricing derivatives. Exploration vs exploitation in learning. Approximate linear programming and Tetris. Lyapunov functions for proving convergence. Viterbi algorithm for decoding, speech recognition, bioinformatics, etc.

Students are welcome to propose other topics, but may have to identify suitable reading material before they are included in the schedule. Actual Schedule: I will fill in this table as we progress through the term. Topics of future lectures are subject to change.

## Publications

If you have problems, please contact the instructor. System representation.

VR lecture 1, sections Richard Baraniuk , speaking on Compressive Signal Processing. This topic is not directly related to dynamic programming, but maybe it could be Value function. Dynamic programming principle. Feedback policies. BS lecture 2 ; VR lecture 1, section 4 , lecture 2, sections DF lecture 1, sections 2, 4, 5 ; BB 1. BS lecture 3.

BB 2; BT 2. BS lecture 4. Efficiency improvements. Infinite horizon problems. Policy iteration. Value iteration applet from David Poole's Interactive Demos. DF lecture 1, section 6 , lecture 2 , lecture 3 ; VR lecture 2, sections , lecture 3. Control, v. This talk will describe how to solve the problems that were posed in Baraniuk's talk from January DF lecture 13 and lecture VR lecture 9 and lecture 10, sections BS lecture Emtiyaz Khan.

Emtiyaz's Kalman filter demo: KFdemo. Stephen Boyd's notes on Kalman filters.

### Navigation Bar

A short note by Emtiyaz on information filters. Optimal Stopping Amit Goyal. HJ: Ivan's slides.

Optimal Stopping: Amit's slides. HJ: BS lecture 7.

Q-factors and Q-learning Stephen Pickett. Jamie Sethian's web page has lots of Fast Marching related material, including an applet demonstrating a continuous version of the travelling salesman problem. Eikonal equation: Josna's slides. Q-learning: Stephen's notes.

Q-learning: BS lecture 24 ; DF lecture 7, section 2 and lecture 8, section 2 ; VR lecture 5 , lecture 6 and lecture 7. Some of David Poole's interactive applets Jacek Kisynski. David Poole's Interactive Demos. Linear Programming: Jonatan's slides. Differential dynamic programming Sang Hoon Yeo. Function approximation: Mark's Matlab code. May require minFunc. Function approximation: Mark's slides.

## Handbook of Learning and Approximate Dynamic Programming

DDP: Sang Hoon's slides. Pegasus: Ken's slides. Jonatan's slides.