Designing Experiments to Test Planning Knowledge about Plan-step Order Constraints
نویسندگان
چکیده
A number of techniques have been developed to effectively extract and generalize planning knowledge based on expert demonstration. In this paper we consider a complementary approach to learning in which we may execute experiments designed to test hypothesized planning knowledge. In particular, we describe an algorithm that automatically generates experiments to test assertions about plan-step ordering, under the assumption that order constraints between two steps are independent of other step orderings. Experimenting with plan-step ordering can help identify asserted ordering constraints that are in fact not necessary, as well as uncover necessary ordering constraints not represented previously. The algorithm consists of three parts: identifying the space of step-ordering hypotheses, efficiently generating ordering tests, and planning experiments that use the tests to identify order constraints that are not currently represented. This method is implemented in the CMAX experiment design module and is part of the POIROT integrated learning system.
منابع مشابه
Designing Experiments to Test and Improve Hypothesized Planning Knowledge Derived from Demonstration
A number of techniques have been developed to effectively extract and generalize planning knowledge based on expert demonstration. In this paper we consider a complementary approach to learning in which experiments are designed to test hypothesized planning knowledge. In particular, we describe an algorithm that automatically generates experiments to test assertions about plan-step ordering. Ex...
متن کاملConstraint-Based Case-Based Planning Using Weighted MAX-SAT
Previous approaches to case-based planning often finds a similar plan case to a new planning problem to adapt to solve the new problem. However, in the case base, there may be some other cases that provide helpful knowledge in building the new solution plan. Likewise, from each existing case there may be only certain parts that can be adapted for solving the new problem. In this paper, we propo...
متن کاملProtein Sequencing Experiment Planning Using Analogy
Experiment design and execution is a central activity in the natural sciences. The SeqER system provides a general architecture for the integration of automated planning techniques with a variety of domain knowledge in order to plan scientific experiments. These planning techniques include rule-based methods and, especially, the use of derivational analogy. Derivational analogy allows planning ...
متن کاملAction Constraints for Planning
Recent progress in the applications of propositional planning systems has led to an impressive speed-up of solution time and an increase in tractable problem size. In part, this improvement stems from the use of domain-dependent knowledge in form of state constraints. In this paper we introduce a diierent class of constraints: action constraints. They express domain-dependent knowledge about th...
متن کاملDefeasible Planning
Planning theory has traditionally made the assumption that the planner begins with all relevant knowledge for solving the problem. Planning agents operating in a complex and dynamic environment cannot make that assumption. They are simultaneously planning agents and epistemic agents, and the pursuit of knowledge is driven by the planning. In particular, the search for a plan initiates the searc...
متن کامل