Download presentation

Presentation is loading. Please wait.

Published byElvis Stetson Modified over 3 years ago

1
The Wumpus World! 2012 级 ACM 班 金汶功

2
Hunt the wumpus!

3
Description Performance measure Environment Actuators Sensors: Stench & Breeze & Glitter & Bump & Scream

4
An Example

6
Reasoning via logic

7
Semantics

8
Models

9
Knowledge base Axioms Current States Sensors Actuators Agent Tell Ask Tell Model checking Answer

10
Efficient Model Checking DPLL Early termination Pure symbol heuristic Unit clause heuristic Component analysis …

11
Drawbacks Model checking is NP-complete Knowledge base may tell nothing.

12
Probabilistic Reasoning

13
Full joint probability distribution P(X, Y) = P(X|Y)P(Y) X: {1,2,3,4} -> {0.1,0.2,0.3,0.4} Y: {a,b} -> {0.4, 0.6} P(X = 2, Y = a) = P(X = 2|Y = a)P(Y = a) The probability of all combination of values

15
Normalization

16
The Wumpus World Aim: calculate the probability that each of the three squares contains a pit.

17
Full joint distribution

18
How likely is it that [1,3] has a pit?

19
Using independence

20
Simplification

22
Finally

23
Bayesian Network

24
Simple Example BurglaryEarthquake Alarm(Bark) John CallsMary Calls P(B).001 P(E).002 BEP(A) Truetrue.95 truefalse.94 falsetrue.29 false.001 BarkP(J) true.90 false.05 BarkP(M) true.70 false.01

25
Specification

26
Conditional Independence

27
Exact Inference

29
P1,3known b P3,1P2,2 P1,3P2,2P3,1b True 1 False1 TrueFalseTrue1 False 0 True 1 FalseTrueFalse1 True0 False 0 P(1,3) 0.2 P(known) P(P3,1) 0.2 P(P2,2) 0.2

30
Approximate Inference Markov Chain Monte Carlo Gibbs Sampling Idea: The long-run fraction of time spent in each state is exactly proportional to its posterior probability.

32
Reference http://zh.wikipedia.org/wiki/Hunt_the_Wumpus http://zh.wikipedia.org/wiki/%E8%B4%9D%E5%8F%B6%E6%9 6%AF%E7%BD%91%E7%BB%9C http://zh.wikipedia.org/wiki/%E8%B4%9D%E5%8F%B6%E6%9 6%AF%E7%BD%91%E7%BB%9C Stuart Russell, Peter Norvig Artificial Intelligence—A Modern Approach 3 rd edition, 2010

Similar presentations

OK

‘In which we introduce a logic that is sufficent for building knowledge- based agents!’

‘In which we introduce a logic that is sufficent for building knowledge- based agents!’

© 2018 SlidePlayer.com Inc.

All rights reserved.

To make this website work, we log user data and share it with processors. To use this website, you must agree to our Privacy Policy, including cookie policy.

Ads by Google

Ppt on beer lambert law Ppt on electronic media in communication Ppt on crash fire tenders Stem and leaf display ppt on tv Ppt on 3d hologram technology Ppt on current account deficit by country Ppt on body language in communication Ppt online viewer php file Ppt on interest rate risk Ppt on campus recruitment system online