Question

How Non-monotonic reasoning deals with incomplete and uncertain model?

subject: Artificial Intelligence  

Answer 1

Reasoning

Reasoning is a process of finding or deriving a valid conclusions using existing facts or avaliable knowledges. Reasoning is an essential factor of artificial intelligence then only a machine can think like a human and and perform tasks.

Non-monotonic Reasoning is a type of Reasoning in Artificial intelligence which deals with incomplete and uncertain models. The perceptions of human in daily life is an example for Non-monotonic reasoning.

Example: The knowledge base of a system contains the given facts below:

• Birds can fly
• Penguins cannot fly
• Kiwi is a bird

So, From here the system can come to the conclusion that Kiwi can fly.

In Non-monotonic the conclusion can be invalidated if more knowledge is added to the knowledge base.

In the previous example, if we add a sentence "Kiwi is a penguin" to the knowledge base,the system will concludes that "Kiwi cannot fly" which invalidates the previous conclusion.

Advantages:

• Non-monotonic reasoning can be used for real-world systems for reasoning like robots.
• We can use assumptions or facts to make valid conclusions.

Disadvantages

• Old conclusions can be invalidated by adding more facts to the knowledge base.
• Cannot be used for proving theorems.