Overall computation cost of AI Production System, Types of Control Strategy

Overall computation cost of AI Production System:

 Rule application cost and control strategy cost together constitutes to the over all computation cost of AI production system.

For an efficient AI system to design there must be balancing between the rule application cost and control strategy.

An efficient AI system involves the usage of the techniques that uses large amount of problem related information without acquiring excessive  control cost.

Types of Control Strategy

There are two types of control strategy they are

1. Irrevocable
2. Tentative
3. Backtracking
4. Graph Search Control

Irrevocable Strategy:

It is a type of control strategy in which once the rule is applied it cannot be retraced.

Here in this, the applicable rule is applied without any provision for reconsideration.

Tentative Strategy:

App rule is applied but provision is made to return later to this point in the computation to apply some other rule. Tentative strategy includes:

Backtracking:

 In this a point is established to which the state of communication can revert to this point.

Graph Search:

 In this, there is a provision that  is made for monitoring the effects of several sequences of rules simultaneously.