Comments (5)
I think the idea was to re-use lookahead
and generally mimic the style used to make the policy gradient algorithms more compact.
I agree that this formulation has more indirection and actual implementations would not look like that.
@mykelk would you consider changing this back to what we had before?
from decisionmaking.
I think it is actually not correct as written now. I think there should be a loop over all s'
around the U'
line. If you fix that, I think it would actually be clear.
from decisionmaking.
I think it's correct as it is. That U'(s)
is just an approximation of the utility function at the next step. The loop tries out different actions and uses one-step lookahead with that utility function as the approximation in the next step. It's different from how it is normally presented, but when I taught it last quarter, this seemed to be easier for the students to understand.
from decisionmaking.
from decisionmaking.
To be concrete about the reason that the DMU version seems clearer - it is recursing through one function, while the new one is recursing through three (forward_search
, lookahead
, and U'
), so it is a bit harder to follow imo
from decisionmaking.
Related Issues (20)
- 4 Parameter Learning. 4.1 Maximum Likelihood Parameter Learning HOT 2
- Norm in SetCategorical HOT 3
- "Backpropagation" in MCTS implementation HOT 2
- Chapter 10, Eq (10.13) HOT 3
- Wording in Chapter 10.6: Mirror Sampling HOT 2
- Wrong reference HOT 1
- Suggestion for line plot in Example 9.8 HOT 6
- Example 22.5 - Incorrect observation used in the second update HOT 1
- Exercise 22.3 - Inconsistent based on Algorithm 22.2 and incorrect observation probabilities HOT 3
- Typo in Algorithm 21.9 HOT 7
- Error in equation D.6 HOT 1
- A small question on page 115, Eq.(6.3) HOT 1
- Caption alg 21.8 HOT 1
- Page 27, Example 2.4 - Equivalence of joint prob. table and tree representation HOT 3
- Equation 7.22 HOT 2
- Missing loop code for algorithms in chapter 5 on descent methods HOT 4
- LightGraphs.jl / topological_sort HOT 4
- Example 4.1 HOT 4
- Example 4.2 HOT 2
- Chapter 8, fig. 8.13 HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from decisionmaking.