Representation Learning for Classical Planning from Partially Observed Traces

Xiao, Zhou; Wan, Hai; Zhuo, Hankui Hankz; Lin, Jinxia; Liu, Yanan

doi:10.48550/arxiv.1907.08352

Cited by 2 publications

(3 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Also, similar to other variants of A* search, our approach is not suitable for problems with a high-dimensional search space. To overcome such limitations, one interesting direction for future work is to extend the proposed approach to planning on general graphs as done by Yang et al (2020); Xiao et al (2019) and with a suitable encoder for high-dimensional spaces (Qureshi et al 2019).…”

Section: Resultsmentioning

confidence: 99%

Path Planning using Neural A* Search

Yonetani,

Taniai,

Barekatain

et al. 2020

Preprint

View full text Add to dashboard Cite

We present Neural A*, a novel data-driven search algorithm for path planning problems. Although data-driven planning has received much attention in recent years, little work has focused on how search-based methods can learn from demonstrations to plan better. In this work, we reformulate a canonical A* search algorithm to be differentiable and couple it with a convolutional encoder to form an end-to-end trainable neural network planner. Neural A* solves a path planning problem by (1) encoding a visual representation of the problem to estimate a movement cost map and (2) performing the A* search on the cost map to output a solution path. By minimizing the difference between the search results and ground-truth paths in demonstrations, the encoder learns to capture a variety of visual planning cues in input images, such as shapes of dead-end obstacles, bypasses, and shortcuts, which makes estimated cost maps informative. Our extensive experiments confirmed that Neural A* (a) outperformed state-of-the-art data-driven planners in terms of the search optimality and efficiency trade-off and (b) predicted realistic pedestrian paths by directly performing a search on raw image inputs.

show abstract

Section: Resultsmentioning

confidence: 99%

Path Planning using Neural A* Search

Yonetani,

Taniai,

Barekatain

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…𝑑𝑜𝑛𝑒 ← true Upon reaching a state from which a plan to the goal exists, the agent stops exploring. For each of its subgoal learners ℓ σ𝑠𝑔 ∈ 𝐿, it attempts to construct sets of preconditions (characterized as a partial fluent state σ𝑝𝑟𝑒 ) from which its policy can consistently achieve the subgoal state σ𝑠𝑔 (see Section 3.4) (lines [28][29][30][31][32][33][34]. If any such precondition sets σ𝑝𝑟𝑒 exist, the agent constructs the operator 𝑜 ★ such that 𝑝𝑟𝑒 (𝑜 ★ ) = σ𝑝𝑟𝑒 , eff (𝑜 ★ ) = σ𝑠𝑔 , and without static fluents (all other variables are unknown once the operator is executed; 𝑠𝑡𝑎𝑡𝑖𝑐 (𝑜 ★ ) = ∅).…”

Section: Learning Operator Policiesmentioning

confidence: 99%

“…The examples consist of successful fluent state and operator pairs corresponding to a sequence of transitions in the symbolic domain [35]. Subsequent work has explored how domains can be learned with partial knowledge of successful traces [2,6], and with neural networks capable of approximating from partial traces [33] and learning models from pixels [4,7].…”

Section: Learning Symbolic Action Modelsmentioning

confidence: 99%

SPOTTER: Extending Symbolic Planning Operators through Targeted Reinforcement Learning

Sarathy¹,

Kasenberg²,

Goel³

et al. 2020

Preprint

View full text Add to dashboard Cite

Symbolic planning models allow decision-making agents to sequence actions in arbitrary ways to achieve a variety of goals in dynamic domains. However, they are typically handcrafted and tend to require precise formulations that are not robust to human error. Reinforcement learning (RL) approaches do not require such models, and instead learn domain dynamics by exploring the environment and collecting rewards. However, RL approaches tend to require millions of episodes of experience and often learn policies that are not easily transferable to other tasks. In this paper, we address one aspect of the open problem of integrating these approaches: how can decision-making agents resolve discrepancies in their symbolic planning models while attempting to accomplish goals? We propose an integrated framework named SPOTTER that uses RL to augment and support ("spot") a planning agent by discovering new operators needed by the agent to accomplish goals that are initially unreachable for the agent. SPOTTER outperforms pure-RL approaches while also discovering transferable symbolic knowledge and does not require supervision, successful plan traces or any a priori knowledge about the missing planning operator.

show abstract

Representation Learning for Classical Planning from Partially Observed Traces

Cited by 2 publications

References 22 publications

Path Planning using Neural A* Search

Path Planning using Neural A* Search

SPOTTER: Extending Symbolic Planning Operators through Targeted Reinforcement Learning

Contact Info

Product

Resources

About