Artificial Intelligence

Symbolic Reasoning under uncertainty

Story so far
We have described techniques for reasoning with a complete, consistent
and unchanging model of the world. But in many problem domains, it is
not possible to create such models. So here we are going to explore
techniques for solving problems with incomplete and uncertain models.

08/12/21 1
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Introduction to Non-monotonic Reasoning

Non monotonic reasoning is one in which the axioms and/or the rules of
inference are extended to make it possible to reason with incomplete
information.
These systems preserve, however, the property that , at any given
moment, a statement is either believed to be true, believed to be false, or
not believed to be either.
Statistical Reasoning : in which the representation is extended to allow
some kind of numeric measure of certainty(rather than true or false) to be
associated with each statement.

08/12/21 2
 Artificial Intelligence

Symbolic Reasoning under uncertainty

At times we need to maintain many parallel belief spaces, each of which
would correspond to the beliefs of one agent.
Such techniques are complicated by the fact that the belief spaces of various
agents, although not identical, are sufficiently similar that it is unacceptably in
efficient to represent them as completely separate knowledge bases.
Conventional reasoning systems, such as FOPL are designed to work with
information that has three important properties.
It is complete with respect to domain of interest.
It is consistent.
The only way it can change is that new facts can be added as they
become available.

08/12/21
All this leads to monotonicity. 3
 Artificial Intelligence

Symbolic Reasoning under uncertainty

If any of these properties is not satisfied, conventional logic based reasoning
systems become inadequate. Non monotonic reasoning systems, are designed
to be able to solve problems in which all of these properties may be missing
Issues to be addressed
How can the knowledge base be extended to allow inferences to be
made on the basis of lack of knowledge as well as on the presence of it?
How can the knowledge base be updated properly when a new fact is
added to the system(or when the old one is removed)?
How can knowledge be used to help resolve conflicts when there are
several in consistent non monotonic inferences that could be drawn?

08/12/21 4
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Logic for monotonic reasoning
Monotonicity is kind of a definition to FOPL, we have to find some
alternative to support non monotonic reasoning.
No single formalization has all the desired properties.
We want to find a formalism that does all of the following things
Define the set of possible worlds that could exist given the facts
that we do have.
Provide a way to say that we prefer to believe in some models
rather than others.
Provide the basis for a practical implementation of this kind of
reasoning.
Corresponds to our intuitions about how this kind of reasoning
08/12/21
works. 5
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Default Reasoning
We use non monotonic reasoning to perform, what is commonly called
Default Reasoning.
We want to draw conclusions based on what is most likely to be true.
Two approaches are
Nonmonotonic Logic
Default Logic
Two common kinds of nonmonotonic reasoning that can be defined in
these logics :
Abduction
Inheritance

08/12/21 6
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Non Monotonic Logic(NML)
It is one in which the language of FOPL is augmented with a modal
operator M, which can be read as “is consistent.”
Vx,y : Related(x,y) ^ M GetAlong(x,y) WillDefend(x,y)
For all x and y, if x and y are related and if the fact that x gets along
with y is consistent with everything else that is believed, then conclude
that x will defend y.
If we are allowing statements in this form, one important issue that must
be resolved if we want our theory to be even semi decidable, we must
decide what “is consistent ” means.

08/12/21 7
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Default Logic
An alternative logic for performing default-based reasoning is Reiter’s
Default Logic(DL) in which a new class of inference rules is introduced.
In this approach, we allow inference rules of this form
A : B
C
The rule should be read as – If A is provable and it is consistent to
assume B, then conclude C

08/12/21 8
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Abduction
Standard logic performs deductions. Given 2 axioms
Vx : A(x)  B(x)
A(C)
We conclude B(C) using deduction
Vx : Measels Spots(x)
 To conclude that if somebody has spots will surely have measels is
incorrect, but it may be the best guess we can make about what is going on.
Deriving conclusions in this way is this another form of default reasoning. We
call this abductive reasoning.

08/12/21 9
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Minimalist Reasoning
We describe methods for saying a very specific and highly useful class
of things that are generally true.
These methods are based on some variant of the idea of a minimal
model.
We will define a model to be minimal if there are no other models in
which fewer things are true.
The idea behind using minimal models as a basis for nonmonotonic
reasoning about the world is the following –
There are many fewer true statements than false ones. If something
is true and relevant it makes sense to assume that it has been entered
into our knowledge base. Therefore, assume that the only true
statements are those that necessarily must be true in order to maintain
08/12/21
the consistency. 10
 Artificial Intelligence

Symbolic Reasoning under uncertainty

The Closed World Assumption
CWA says that the only objects that satisfy any predicate P are those
that must.
Eg. A company’s employee database.
Airline example
Although the CWA is both simple & powerful, it can fail to produce an
appropriate answer for either of the two reasons.
The assumptions are not always true in the world; some parts of the
world are not realistically “closable”. - unrevealed facts in murder case
It is a purely syntactic reasoning process. Thus, the result depend on the
form of assertions that are provided.

08/12/21 11
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Augment a Problem- Solver
How to write a program that solves problems using axioms.
Even uncertain knowledge can be solved using forward & backward
reasoning.
As a result there are 2 approaches to this kind of problem solving
Reason forward from what is known. Treat non monotonically
derivable conclusions the same way monotonically derivable ones are
handled.
Everything is same as monotonic except that here we have forward
chaining rules but with UNLESS clause.
Reason backward to determine whether some expressions P is true?

08/12/21 12
 Artificial Intelligence

Symbolic Reasoning under uncertainty

Augment a Problem- Solver
Reason backward to determine whether some expressions P is true?
(or perhaps to find a bindings for its variables that make it true) Non-
monotonic reasoning systems that support this kind of reasoning may
do either or both of the following
Allow default(unless) clauses in backward ruling.
Support a kind of debate in which an attempt is made to
construct arguments both in favor of P and opposed to it.

08/12/21 13