+ All Categories
Home > Documents > 4/29: Conditional Planning No Final. Instead we will have a last homework Midterm to be returned...

4/29: Conditional Planning No Final. Instead we will have a last homework Midterm to be returned...

Date post: 15-Jan-2016
Category:
View: 214 times
Download: 0 times
Share this document with a friend
86
4/29: Conditional Planning No Final. Instead we will have a last homework Midterm to be returned Thursday; Homework reached Hanoi Extra class on Thursday?
Transcript
Page 1: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

4/29: Conditional Planning

No Final. Instead we will have a last homework

Midterm to be returned Thursday; Homework reached Hanoi

Extra class on Thursday?

Page 2: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

ConformantHeuristics

Page 3: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Sensing: General observations Sensing can be thought in terms of

Speicific state variables whose values can be found OR sensing actions that evaluate truth of some boolean formula over the state variables.

Sense(p) ; Sense(pV(q&r)) A general action may have both causative effects and sensing effects

Sensing effect changes the agent’s knowledge, and not the world Causative effect changes the world (and may give certain knowledge to the agent) A pure sensing action only has sensing effects; a pure causative action only has causative

effects. When applied to a belief state, the sensing effects of an action wind up reducing the

cardinality of that belief state basically by removing all states that are not consistent with the sensed effects

Sensing actions introduce branches into the plans If you apply Sense-A? to a belief state B, you get a partition of B: BA and B~A

You will have to make a plan for both the branches. And/Or search in the space of belief states

Page 4: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

First we will see a model where observability is in terms of state variablesNext we shall see a model where observability can be in terms of formulas [Bonet&Geffner]

Page 5: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Modeling observability in terms of observation actions that give values of single state variables

If a state variable pIs in B, then there is some action Ap thatCan sense whether p is true or false

Page 6: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 7: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 8: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 9: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 10: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

A Simple Progression Algorithm in the presence of pure sensing actions

Call the procedure Plan(BI,G,nil) where Procedure Plan(B,G,P)

If G is satisfied in all states of B, then return P Non-deterministically choose:

I. Non-deterministically choose a causative action a that is applicable in B. Return Plan(a(B),G,P+a)

II. Non-deterministically choose a sensing action s that senses a formula f (could be a single state variable)

Let p’ = Plan(Bf,G,nil); p’’=Plan(B~f,G,nil) /*Bf is the set of states of B in which f is true */

Return P+(s?:p’;p’’)

If we always pick I and never do II then we will produce conformantPlans (if we succeed).

Page 11: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Remarks on the progression with sensing actions

Progression is implicitly finding an AND subtree of an AND/OR Graph If we look for AND subgraphs, we can represent looping

plans. The amount of sensing done in the eventual solution

plan is controlled by how often we pick step I vs. step II (if we always pick I, we get conformant solutions). Progression is as clue-less as to whether to do sensing and

which sensing to do, as it is about which causative action to apply Need heuristic support

Page 12: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Geffner’s model

Page 13: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 14: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 15: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 16: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 17: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 18: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 19: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 20: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

5/1: Conditional Planning (contd)

Midterms returnedMonty Python on Conformant PlanningNext class *REQUIRED READING* assignedCan we meet sometime on Monday (instead of reading class?)

Page 21: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Very simple Example

A1 p=>r,~pA2 ~p=>r,p

A3 r=>g

O5 observe(p)

Problem: Init: don’t know p Goal: g

Plan: O5:p?[A1A3][A2A3]

O5:p?

A1

A3

A2

A3

YN

Page 22: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 23: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 24: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 25: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

nuPDDL—not yet a standard… define (domain d) ... (:predicates (P1) (P2)

(P3) (P4) (P5) (P6)) ...) (define (problem p)

.... (:init (and (P1) (oneof (and (P2) (P3)) (P4)) (unknown (P5)))) ....)

define (problem p) .... (:effect (and (P1) (oneof (and (P2) (P3)) (P4)) (unknown (P5)))) ....)

(:observation wall_north - boolean :parameters () (iff (= wall_north 1) (or (= (robot_y) north) (= (robot_x) west))))

;.... (:observation wall_east - boolean :parameters () (imply (= wall_east 0) (= (robot_x) west)) (imply (= wall_east 1) (true))

:weakgoal : it is required that the plan may reach the goal. :stronggoal : it is required that every execution of the plan reaches the goal. :strongcyclicgoal : it is required that every execution of the plan either reaches the goal, or at least always has a chance to do it :postronggoal : it is required that every execution of the plan reaches the goal, using only the observations described in the domain :conformantgoal : it is required that every execution of the plan reaches the goal, without ever observing :ctlgoal : it is required that the CTL formula expressed as a goal is satisfied throughout every possible execution of the plan. Some

examples of typical extended goals follow: Do Reach p (``strong goal''): (af p) Try Reach p (``weak goal''): (ef p) Keep Trying Reach p (``strong cyclic goal''): (aw (ef p) p)

http://sra.itc.it/tools/mbp/NuPDDL.html

Page 26: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

A Simple Progression Algorithm in the presence of pure sensing actions

Call the procedure Plan(BI,G,nil) where Procedure Plan(B,G,P)

If B is a subset of BG (or any B’ in P that is marked “solved”) return P (propagate “solve” marking upwards)

Non-deterministically choose: I. Non-deterministically choose a causative action a that is applicable in B.

Return Plan(a(B),G,P+a) II. Non-deterministically choose a sensing action s that senses a formula f

(could be a single state variable) Let p’ = Plan(Bf,G,nil); p’’=Plan(B~f,G,nil)

/*Bf is the set of states of B in which f is true */ Return P+(s?:p’;p’’)

If we always pick I and never do II then we will produce conformantPlans (if we succeed).

Page 27: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Remarks on the progression with sensing actions

Progression is implicitly finding an AND subgraph of an AND/OR Graph

The amount of sensing done in the eventual solution plan is controlled by how often we pick step I vs. step II (if we always pick I, we get conformant solutions). Progression is as clue-less as to whether to do sensing and

which sensing to do, as it is about which causative action to apply Need heuristic support

Page 28: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Cost models of conditional plans

The execution cost of a conditional plan isCost of O5 +[Prob(p=T)* {cost of A1 + A3} +

Prob(p=F)*{cost of A2 +A3} ] Can take max(cost A1+A3; cost A2+A3

) The planning cost of a conditional

plan is however is proportional to the total size of the plan (num actions)

O5:p?

A1

A3

A2

YN

O5:p?

A1A2

YN

Need to estimate cost of leaf belief states

Page 29: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 30: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 31: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 32: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 33: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 34: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 35: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 36: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 37: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Similar processing can be done for regression (PO planning is nothing but least-committed regression planning)

Page 38: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Sensing: General observations Sensing can be thought in terms of

Speicific state variables whose values can be found OR sensing actions (with preconditions and causative effects) that evaluate truth

of some boolean formula over the state variables. Sense(p) ; Sense(pV(q&r))

A general action may have both causative effects and sensing effects Sensing effect changes the agent’s knowledge, and not the world Causative effect changes the world (and may give certain knowledge to the agent) A pure sensing action only has sensing effects; a pure causative action only has

causative effects. The recent work on conditional planning has considered mostly simplistic sensing

actions that have no preconditions and only have pure sensing effects. When applied to a belief state, the sensing effects of an action wind up

reducing the cardinality of that belief state basically by removing all states that are not consistent with the sensed effects

Sensing actions introduce branches into the plans If you apply Sense-A? to a belief state B, you get a partition of B: BA and B~A

You will have to make a plan for both the branches. And/Or search in the space of belief states

Page 39: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Sensing: More things under the mat

Sensing extends the notion of goals too. Check if Rao is awake vs. Wake up Rao

Presents some tricky issues in terms of goal satisfaction…!

Handling quantified effects and preconditions in the presence of sensing actions Rm* can satisfy the effect forall files remove(file); without KNOWING what are the files in the directory!

Sensing actions can have preconditions (as well as other causative effects)

The problem of OVER-SENSING (Sort of like the initial driver; also Sphexishness) [XII/Puccini project] Handling over-sensing using local-closedworld assumptions

Listing a file doesn’t destroy your knowledge about the size of a file; but compressing it does. If you don’t recognize it, you will always be checking the size of the file after each and every action

A general action may have both causative effects and sensing effects Sensing effect changes the agent’s knowledge, and not the world Causative effect changes the world (and may give certain knowledge to the agent) A pure sensing action only has sensing effects; a pure causative action only has causative effects.

The recent work on conditional planning has considered mostly simplistic sensing actions that have no preconditions and only have pure sensing effects.

Sensing has cost!

Page 40: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Sensing: Limited Contingency planning

In many real-world scenarios, having a plan that works in all contingencies is too hard An idea is to make a plan for some of the contingencies; and

monitor/Replan as necessary. Qn: What contingencies should we plan for?

The ones that are most likely to occur…(need likelihoods) Qn: What do we do if an unexpected contingency arises?

Monitor (the observable parts of the world) When it goes out of expected world, replan starting from that state.

Page 41: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 42: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

“Triangle Tables”

Page 43: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 44: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 45: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 46: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 47: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 48: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 49: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 50: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 51: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 52: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

NOT used beyond this point

Page 53: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 54: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 55: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 56: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 57: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 58: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 59: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 60: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 61: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 62: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 63: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 64: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 65: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 66: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 67: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 68: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 69: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 70: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 71: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 72: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 73: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 74: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 75: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 76: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 77: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 78: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 79: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 80: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 81: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 82: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 83: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 84: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 85: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.
Page 86: 4/29: Conditional Planning  No Final. Instead we will have a last homework  Midterm to be returned Thursday; Homework reached Hanoi  Extra class on.

Recommended