Learning complex action models with quantifiers and logical implications

Zhuo, Hankz Hankui; Yang, Qiang; Hu, Derek Hao; Li, Lei

doi:10.1016/j.artint.2010.09.007

Cited by 64 publications

(56 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In general, creating a domain for a real-world problem is a hard task. This has prompted the implementation of algorithms that learn such domains from traces of plan executions (Zhuo et al 2010;Nejati et al 2009). However, the application of such methods to robotic problems remains to be demonstrated (preliminary results can be found in Philippsen et al 2009).…”

Section: On the Representationmentioning

confidence: 99%

Autonomous construction using scarce resources in unknown environments

2012

View full text Add to dashboard Cite

The goal of creating machines that autonomously perform useful work in a safe, robust and intelligent manner continues to motivate robotics research. Achieving this autonomy requires capabilities for understanding the environment, physically interacting with it, predicting the outcomes of actions and reasoning with this knowledge. Such intelligent physical interaction was at the centre of early robotic investigations and remains an open topic.In this paper, we build on the fruit of decades of research to explore further this question in the context of autonomous construction in unknown environments with scarce resources. Our scenario involves a miniature mobile robot that autonomously maps an environment and uses cubes to bridge ditches and build vertical structures according to high-level goals given by a human.Based on a "real but contrived" experimental design, our results encompass practical insights for future applications that also need to integrate complex behaviours under hardware constraints, and shed light on the broader question of

show abstract

Section: On the Representationmentioning

confidence: 99%

Autonomous construction using scarce resources in unknown environments

2012

View full text Add to dashboard Cite

show abstract

“…REX-D can address partially observable domains if the learner and the planner used are able to learn them. Other approaches have been proposed that learn complex models in partially observable domains [26]. However, we use Pasula et al's learner because its generalization capabilities and its ability to tackle uncertain action effects are better suited than partial observability learners to the tasks considered in the present study.…”

Section: Learnermentioning

confidence: 99%

Relational reinforcement learning with guided demonstrations

Martínez

Alenyà

Torras

2017

Artificial Intelligence

View full text Add to dashboard Cite

Model-based reinforcement learning is a powerful paradigm for learning tasks in robotics. However, in-depth exploration is usually required and the actions have to be known in advance. Thus, we propose a novel algorithm that integrates the option of requesting teacher demonstrations to learn new domains with fewer action executions and no previous knowledge. Demonstrations allow new actions to be learned and they greatly reduce the amount of exploration required, but they are only requested when they are expected to yield a significant improvement because the teacher's time is considered to be more valuable than the robot's time. Moreover, selecting the appropriate action to demonstrate is not an easy task, and thus some guidance is provided to the teacher. The rule-based model is analyzed to determine the parts of the state that may be incomplete, and to provide the teacher with a set of possible problems for which a demonstration is needed. Rule analysis is also used to find better alternative models and to complete subgoals before requesting help, thereby minimizing the number of requested demonstrations. These improvements were demonstrated in a set of experiments, which included domains from the international planning competition and a robotic task. Adding teacher demonstrations and rule analysis reduced the amount of exploration required by up to 60% in some domains, and improved the success ratio by 35% in other domains.

show abstract

“…Note that we build the action models based on the study of the data communication in P2P networks. On the other hand, we need to point out that there are some other ways which are helpful to generate the definition of action models, such as ARMS and LAMP algorithms which can automatically discover action models from a set of successful observed plans [Yang et al 2007;Zhuo et al 2010]. There are at least two different types of optimization in P2P networks.…”

Section: Testing Domainsmentioning

confidence: 99%

A SAT-based approach to cost-sensitive temporally expressive planning

Lü

Huang

Chen

et al. 2013

ACM Trans. Intell. Syst. Technol.

View full text Add to dashboard Cite

Complex features, such as temporal dependencies and numerical cost constraints, are hallmarks of real-world planning problems. In this paper, we consider the challenging problem of cost-sensitive temporally expressive (CSTE) planning, which requires concurrency of durative actions and optimization of action costs. We first propose a scheme to translate a CSTE planning problem to a minimum cost (MinCost) satisfiability (SAT) problem and to integrate with a relaxed parallel planning semantics for handling true temporal expressiveness. Our scheme finds solution plans that optimize temporal makespan, and also minimize total action costs at the optimal makespan. We propose two approaches for solving MinCost SAT. The first is based on a transformation of a MinCost SAT problem to a weighted partial Max-SAT (WPMax-SAT), and the second, called BB-CDCL, is an integration of the branch-and-bound technique and the conflict driven clause learning (CDCL) method. We also develop a CSTE customized variable branching scheme for BB-CDCL which can significantly improve the search efficiency. Our experiments on the existing CSTE benchmark domains show that our planner compares favorably to the state-of-the-art temporally expressive planners in both efficiency and quality.

show abstract

Learning complex action models with quantifiers and logical implications

Cited by 64 publications

References 30 publications

Autonomous construction using scarce resources in unknown environments

Autonomous construction using scarce resources in unknown environments

Relational reinforcement learning with guided demonstrations

A SAT-based approach to cost-sensitive temporally expressive planning

Contact Info

Product

Resources

About