LAO*: A heuristic search algorithm that finds solutions with loops

By Eric A. Hansen, Shlomo Zilberstein. Artificial Intelligence, 2001

510 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Information Systems, Software

View Paper: https://www.semanticscholar.org/paper/20f0d7d9369ae0fc8ba0a9713738d993d30d7e80

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Hector Geffner
2. Blai Bonet
3. Malik Ghallab
4. Paolo Traverso
5. Joelle Pineau
6. Dana S. Nau
7. Miquel Ramírez
8. R. Zhou

In Information Systems

1. Hector Geffner
2. Malik Ghallab
3. Blai Bonet
4. Paolo Traverso
5. Aijun Bai
6. Dana S. Nau
7. R. Zhou
8. F. Ingrand

In Software

1. Hector Geffner
2. Blai Bonet
3. Malik Ghallab
4. Aijun Bai
5. R. Zhou
6. F. Ingrand
7. Martijn van Otterlo
8. Marco A. Wiering

Citations in the corpus

• Anytime Heuristic Search (2011, methodology): “…other search algorithms besides A*, including memoryefficient versions of A* such as IDA* and RBFS (Korf, 1993), as well as Learning RealTime A* (LRTA*) (Shimbo & Ishida, 2003), and heuristic search algorithms for AND/OR graphs (Chakrabarti, Ghosh, & DeSarkar, 1988; Hansen & Zilberstein, 2001).”
• Reinforcement Learning and Markov Decision Processes (2012, methodology): “Similar to the previous methods the LAO* algorithm Hansen and Zilberstein (2001) also alternates between an expansion phase and a policy generation phase.”
• Plan Recognition as Planning (2009, background): “In Goal Mdps M there is no similar closed–form bound, although it can be computed at some expense (Bertsekas, 1995; Hansen and Zilberstein, 2001).”
• Online Planning Algorithms for POMDPs (2008, methodology): “These algorithms could always be modified to handle generic graph structures by using a technique proposed in the LAO* algorithm (Hansen & Zilberstein, 2001) to handle cycles.”
• Online Planning for Large Markov Decision Processes with Hierarchical Decomposition (2015, methodology): “BACKGROUND In this section, we briefly review the MDP model [Puterman”
• Labeled RTDP: Improving the Convergence of Real-Time Dynamic Programming (2003, methodology): “In stochastic shortest path models such as M1– M7, there is no similar closed-form bound, although such bound can be computed at some expense (Bertsekas 1995; Hansen & Zilberstein 2001).”
• Deliberation for autonomous robots: A survey (2017, methodology): “That framework can be implemented in di ff erent ways, e.g., with a best-first search, as in AO*, LAO* [95] and their extensions, with a depth-first iterative deepening search, as in LDFS [27], or with a random walk search, as in RTDP [13], LRTDP [26] and their extensions.”
• Automated Planning and Acting (2016, methodology): “Hansen and Zilberstein [254] developed the LAO∗ algorithm as an extension of AO∗ of Nilsson [460].”
• An Algorithm Better than AO*? (2005, methodology): “2A similar observation appears in (Hansen & Zilberstein 2001).”
• Markovian State and Action Abstractions for MDPs via Hierarchical MCTS (2016, methodology): “An online planning algorithm finds the best action for the current state by exploring an expectimax search tree starting from the current state [Barto et al., 1995; Hansen and Zilberstein, 2001; Kocsis and Szepesvári, 2006].”
• Action Schema Networks: Generalised Policies with Deep Learning (2017, methodology): “We compare ASNet against three probabilistic planners based on heuristic search: LRTDP (Bonet and Geffner 2003), iLAO* (Hansen and Zilberstein 2001) and SSiPP (Trevizan and Veloso 2014).”