var bibbase_data = {"data":"\"Loading..\"\n\n
\n\n \n\n \n\n \n \n\n \n\n \n \n\n \n\n \n
\n generated by\n \n \"bibbase.org\"\n\n \n
\n \n\n
\n\n \n\n\n
\n\n Excellent! Next you can\n create a new website with this list, or\n embed it in an existing web page by copying & pasting\n any of the following snippets.\n\n
\n JavaScript\n (easiest)\n
\n \n <script src=\"https://bibbase.org/show?bib=https://icaps22.icaps-conference.org/dc/dc-papers.bib&theme=default&fullnames=1&hidemenu=1&jsonp=1&jsonp=1\"></script>\n \n
\n\n PHP\n
\n \n <?php\n $contents = file_get_contents(\"https://bibbase.org/show?bib=https://icaps22.icaps-conference.org/dc/dc-papers.bib&theme=default&fullnames=1&hidemenu=1&jsonp=1\");\n print_r($contents);\n ?>\n \n
\n\n iFrame\n (not recommended)\n
\n \n <iframe src=\"https://bibbase.org/show?bib=https://icaps22.icaps-conference.org/dc/dc-papers.bib&theme=default&fullnames=1&hidemenu=1&jsonp=1\"></iframe>\n \n
\n\n

\n For more details see the documention.\n

\n
\n
\n\n
\n\n This is a preview! To use this list on your own web site\n or create a new web site from it,\n create a free account. The file will be added\n and you will be able to edit it in the File Manager.\n We will show you instructions once you've created your account.\n
\n\n
\n\n

To the site owner:

\n\n

Action required! Mendeley is changing its\n API. In order to keep using Mendeley with BibBase past April\n 14th, you need to:\n

    \n
  1. renew the authorization for BibBase on Mendeley, and
    2. \n
  2. update the BibBase URL\n in your page the same way you did when you initially set up\n this page.\n
    3. \n
\n

\n\n

\n \n \n Fix it now\n

\n
\n\n
\n\n\n
\n \n \n
\n
\n  \n 2022\n \n \n (14)\n \n \n
\n
\n \n \n
\n \n\n \n \n \n \n \n \n Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022).\n \n \n \n \n\n\n \n Pascal Bercher; and Sara Bernardini.,\n editors.\n \n\n\n \n\n\n\n 2022.\n \n\n\n\n
\n\n\n\n \n \n \"Proceedings webpage\n  \n \n \n \"Proceedings proceedings\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 12 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@Proceedings{ICAPS-DC-2022,\n  title           = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year            = {2022},\n  editor          = {Pascal Bercher and Sara Bernardini},\n  abstract        = {This is the proceedings of the ICAPS Doctoral Consortium 2022.},\n  url_webpage     = {https://icaps22.icaps-conference.org/dc-2022},\n  url_proceedings = {https://icaps22.icaps-conference.org/dc/DC-Proceedings-2022.pdf}\n}\n\n
\n
\n\n\n
\n This is the proceedings of the ICAPS Doctoral Consortium 2022.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n A Generalization of Automated Planning Using Dynamically Estimated Action Models.\n \n \n \n \n\n\n \n Eyal Weiss.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 1–3, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"A paper\n  \n \n \n \"A presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 9 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-01,\n  author           = {Eyal Weiss},\n  title            = {A Generalization of Automated Planning Using Dynamically Estimated Action Models},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {1--3},\n  abstract         = {Representing real-world planning problems is a major open subject. Standard planning modeling languages are fully declarative, making it challenging to use them for expressing complex mathematical functions, that are often required for describing the effects of actions. Recent approaches turn to external sources of information, such as simulators or black-box modules, to overcome such modeling limitations. This paper proposes a novel approach to represent and solve planning problems, by starting with partial declarative action models and incrementally refining them during planning by invoking domain-specific external modules. Since these might be computationally expensive, we provide the planner the ability to trade-off modeling uncertainty against computation time, to meet target plan accuracy. Results that were obtained for planning with dynamic estimation of action costs are sketched, and planned work, together with open challenges, are further detailed.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_364.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=J3JKlX8wuVg&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV&index=9}\n}\n\n
\n
\n\n\n
\n Representing real-world planning problems is a major open subject. Standard planning modeling languages are fully declarative, making it challenging to use them for expressing complex mathematical functions, that are often required for describing the effects of actions. Recent approaches turn to external sources of information, such as simulators or black-box modules, to overcome such modeling limitations. This paper proposes a novel approach to represent and solve planning problems, by starting with partial declarative action models and incrementally refining them during planning by invoking domain-specific external modules. Since these might be computationally expensive, we provide the planner the ability to trade-off modeling uncertainty against computation time, to meet target plan accuracy. Results that were obtained for planning with dynamic estimation of action costs are sketched, and planned work, together with open challenges, are further detailed.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Action Model Learning based on Grammar Induction.\n \n \n \n \n\n\n \n Maxence Grand.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 4–8, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Action paper\n  \n \n \n \"Action presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 8 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-02,\n  author           = {Maxence Grand},\n  title            = {Action Model Learning based on Grammar Induction},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {4--8},\n  abstract         = {This paper presents a novel approach to learn PDDL domain called AMLSI (Action Model Learning with State machine Interaction) based on grammar induction. AMLSI learns with no prior knowledge from a training dataset made up of action sequences built by random walks and by observing state transitions. The domain learnt is accurate enough to be used without human proofreading in a planner even with very highly partial and noisy observations. Thus AMLSI takles a key issue for domain learning that is the ability to plan with the learned domains. It often happens that small learning errors lead to domains that are unusable for planning. AMLSI contribution is to learn domains from partial and noisy observations with sufficient accuracy to allow planners to solve new problems. Also, this paper presents an incremental and a temporal extension.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_356.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=TbHLkyoZ2sQ&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV&index=5}\n}\n\n
\n
\n\n\n
\n This paper presents a novel approach to learn PDDL domain called AMLSI (Action Model Learning with State machine Interaction) based on grammar induction. AMLSI learns with no prior knowledge from a training dataset made up of action sequences built by random walks and by observing state transitions. The domain learnt is accurate enough to be used without human proofreading in a planner even with very highly partial and noisy observations. Thus AMLSI takles a key issue for domain learning that is the ability to plan with the learned domains. It often happens that small learning errors lead to domains that are unusable for planning. AMLSI contribution is to learn domains from partial and noisy observations with sufficient accuracy to allow planners to solve new problems. Also, this paper presents an incremental and a temporal extension.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Application of Neurosymbolic AI to Sequential Decision Making.\n \n \n \n \n\n\n \n Carlos Núñez-Molina.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 9–12, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Application paper\n  \n \n \n \"Application presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 6 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-03,\n  author           = {Carlos Núñez-Molina},\n  title            = {Application of Neurosymbolic AI to Sequential Decision Making},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {9--12},\n  abstract         = {In the history of AI, two main paradigms have been applied to solve Sequential Decision Making (SDM) problems: Automated Planning (AP) and Reinforcement Learning (RL). Among the many proposals to unify both fields, the one known as neurosymbolic AI has recently attracted great attention. It combines the Deep Neural Networks characteristic of modern RL with the symbolic representations typical of AP. The main goal of this PhD is to progress the state of the art in neurosymbolic AI for SDM. To do so, three different lines of research have been proposed. In the first one, I will perform a study of the literature and summarize my findings into a review. In the second one, I will extend my previous work (Núñez-Molina et al. 2021), which combined Deep Q-Learning with Classical Planning to improve planning performance, with the ability to manage non-determinism and will apply it to a real logistics problem. Finally, in the third line of research, I will develop a method for generating planning problems and leverage it to learn HTN domains without expert traces and to study the properties of planning domains.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_350.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=YtwYa2ENOI8&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV&index=11}\n}\n\n
\n
\n\n\n
\n In the history of AI, two main paradigms have been applied to solve Sequential Decision Making (SDM) problems: Automated Planning (AP) and Reinforcement Learning (RL). Among the many proposals to unify both fields, the one known as neurosymbolic AI has recently attracted great attention. It combines the Deep Neural Networks characteristic of modern RL with the symbolic representations typical of AP. The main goal of this PhD is to progress the state of the art in neurosymbolic AI for SDM. To do so, three different lines of research have been proposed. In the first one, I will perform a study of the literature and summarize my findings into a review. In the second one, I will extend my previous work (Núñez-Molina et al. 2021), which combined Deep Q-Learning with Classical Planning to improve planning performance, with the ability to manage non-determinism and will apply it to a real logistics problem. Finally, in the third line of research, I will develop a method for generating planning problems and leverage it to learn HTN domains without expert traces and to study the properties of planning domains.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Counter-Example Based Planning.\n \n \n \n \n\n\n \n Xiaodi Zhang.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 14–17, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Counter-Example paper\n  \n \n \n \"Counter-Example presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 2 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-04,\n  author           = {Xiaodi Zhang},\n  title            = {Counter-Example Based Planning},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {14--17},\n  abstract         = {CPCES is one of conformant planning problem solvers. It continuously searches candidate plans and counter-examples until finding a valid plan or no solution. The goal of my Ph.D. project includes improving its efficiency, making it compatible with more classical planners, and using it to solve contingent planning problems.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_358.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=wmYwfbNuQJQ&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV&index=1}\n}\n\n
\n
\n\n\n
\n CPCES is one of conformant planning problem solvers. It continuously searches candidate plans and counter-examples until finding a valid plan or no solution. The goal of my Ph.D. project includes improving its efficiency, making it compatible with more classical planners, and using it to solve contingent planning problems.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Data Efficient Paradigms for Personalized Assessment of Taskable AI Systems.\n \n \n \n \n\n\n \n Pulkit Verma.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 18–22, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Data paper\n  \n \n \n \"Data presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 15 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-05,\n  author           = {Pulkit Verma},\n  title            = {Data Efficient Paradigms for Personalized Assessment of Taskable AI Systems},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {18--22},\n  abstract         = {The vast diversity of internal designs of taskable black-box AI systems and their nuanced zones of safe functionality make it difficult for a layperson to use them without unintended side effects. The focus of my dissertation is to develop algorithms and requirements of interpretability that would enable a user to assess and understand the limits of an AI system’s safe operability. We develop a personalized AI assessment module that lets an AI system execute instruction sequences in simulators and answer the queries about its execution of sequences of actions. Our results show that such a primitive query-response capability is sufficient to efficiently derive a user-interpretable model of the system’s capabilities in fully observable, and deterministic settings.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_363.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=7jwlOX-0qrE&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n\n
\n
\n\n\n
\n The vast diversity of internal designs of taskable black-box AI systems and their nuanced zones of safe functionality make it difficult for a layperson to use them without unintended side effects. The focus of my dissertation is to develop algorithms and requirements of interpretability that would enable a user to assess and understand the limits of an AI system’s safe operability. We develop a personalized AI assessment module that lets an AI system execute instruction sequences in simulators and answer the queries about its execution of sequences of actions. Our results show that such a primitive query-response capability is sufficient to efficiently derive a user-interpretable model of the system’s capabilities in fully observable, and deterministic settings.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Domain Specific Situated Planning.\n \n \n \n \n\n\n \n Devin Thomas.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 23–26, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Domain paper\n  \n \n \n \"Domain presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 3 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-06,\n  author           = {Devin Thomas},\n  title            = {Domain Specific Situated Planning},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {23--26},\n  abstract         = {There has been much recent work on finding paths in grid maps among moving obstacles. However, in addition to assuming complete omniscience regarding the map and the obstacles’ trajectories, previous work has also assumed that time stands still while the agent plans. My dissertation addresses situated pathfinding, in which time passes and the obstacles continue to move while the agent plans. I will study situated planning in three domains: Grid pathfinding among moving obstacles, orienteering and opportunistic science.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_362.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=IM8MkG7aejg&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n\n
\n
\n\n\n
\n There has been much recent work on finding paths in grid maps among moving obstacles. However, in addition to assuming complete omniscience regarding the map and the obstacles’ trajectories, previous work has also assumed that time stands still while the agent plans. My dissertation addresses situated pathfinding, in which time passes and the obstacles continue to move while the agent plans. I will study situated planning in three domains: Grid pathfinding among moving obstacles, orienteering and opportunistic science.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Domain-Independent Heuristics in Probabilistic Planning.\n \n \n \n \n\n\n \n Thorsten Klößner.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 27–31, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Domain-Independent paper\n  \n \n \n \"Domain-Independent presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 9 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-07,\n  author           = {Thorsten Kl{\\"o}{\\ss}ner},\n  title            = {Domain-Independent Heuristics in Probabilistic Planning},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {27--31},\n  abstract         = {It has been almost two decades since MDP heuristic search algorithms have been developed. These algorithms guarantee to find an optimal policy for the initial state for several optimization objectives without necessarily expanding the entire state space, if provided with a heuristic that provides optimistic state value estimates. While a large and diverse set of such domain-independent heuristic families is available in classical planning, the same cannot be said about probabilistic planning. So far, except for the particular case of occupation measure heuristics for (constrained) Stochastic Shortest Path Problems, most of the attempts at constructing heuristics pursue the very simple approach of using a classical heuristic on the all-outcomes determinization of the planning problem, in which the probabilistic effect of an action can be chosen at will. Because this approach is agnostic to the uncertainty in the underlying problem, these heuristics are often not very informative. In this thesis, we will investigate heuristics for probabilistic planning which are formulated on the underlying probabilistic model directly instead of delegating to a classical heuristic on the determinization. To this end, we mainly focus on abstraction heuristics, in particular Pattern Database heuristics and Merge-and-Shrink heuristics.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_360.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=a9gbeu4R2y8&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n\n
\n
\n\n\n
\n It has been almost two decades since MDP heuristic search algorithms have been developed. These algorithms guarantee to find an optimal policy for the initial state for several optimization objectives without necessarily expanding the entire state space, if provided with a heuristic that provides optimistic state value estimates. While a large and diverse set of such domain-independent heuristic families is available in classical planning, the same cannot be said about probabilistic planning. So far, except for the particular case of occupation measure heuristics for (constrained) Stochastic Shortest Path Problems, most of the attempts at constructing heuristics pursue the very simple approach of using a classical heuristic on the all-outcomes determinization of the planning problem, in which the probabilistic effect of an action can be chosen at will. Because this approach is agnostic to the uncertainty in the underlying problem, these heuristics are often not very informative. In this thesis, we will investigate heuristics for probabilistic planning which are formulated on the underlying probabilistic model directly instead of delegating to a classical heuristic on the determinization. To this end, we mainly focus on abstraction heuristics, in particular Pattern Database heuristics and Merge-and-Shrink heuristics.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Learning Hierarchical Abstractions for Efficient Taskable Robots.\n \n \n \n \n\n\n \n Naman Shah.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 32–35, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Learning paper\n  \n \n \n \"Learning presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 5 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-08,\n  author           = {Naman Shah},\n  title            = {Learning Hierarchical Abstractions for Efficient Taskable Robots},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {32--35},\n  abstract         = {Although state-of-the-art hierarchical robot planning algorithms allow robots to efficiently compute long-horizon motion plans for achieving user desired tasks, these methods typically rely upon environment-dependent state and action abstractions that need to be hand-designed by experts. On the other hand, non-hierarchical robot planning approaches fail to compute solutions for complex tasks that require reasoning over a long horizon. My research addresses these problems by proposing an approach for learning abstractions and developing hierarchical planners that efficiently use learned abstractions to boost robot planning performance while providing strong guarantees of reliability.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_361.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=jaQqbSdIqjI&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n\n
\n
\n\n\n
\n Although state-of-the-art hierarchical robot planning algorithms allow robots to efficiently compute long-horizon motion plans for achieving user desired tasks, these methods typically rely upon environment-dependent state and action abstractions that need to be hand-designed by experts. On the other hand, non-hierarchical robot planning approaches fail to compute solutions for complex tasks that require reasoning over a long horizon. My research addresses these problems by proposing an approach for learning abstractions and developing hierarchical planners that efficiently use learned abstractions to boost robot planning performance while providing strong guarantees of reliability.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Modeling Assistance for AI Planning From the Perspective of Model Reconciliation.\n \n \n \n \n\n\n \n Songtuan Lin.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 36–40, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Modeling paper\n  \n \n \n \"Modeling presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 11 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-09,\n  author           = {Songtuan Lin},\n  title            = {Modeling Assistance for AI Planning From the Perspective of Model Reconciliation},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {36--40},\n  abstract         = {Providing modeling assistance to domain modelers is a prominent challenge in incorporating humans into planning processes. Many efforts have been devoted to this direction in classical planning, however, only few works have been done in hierarchical planning. In this thesis, we will study a methodology for providing modeling assistance in HTN planning, which is the most commonly used hierarchical planning framework. Particularly, we will address two bottleneck problems for this purpose, namely domain model validation and domain model refinements. For the former one, we propose an approach based on plan verification, and for the latter, we view it as a model reconciliation problem and will study a novel approach for solving it.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_359.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=MUYl845Dy4I&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n\n
\n
\n\n\n
\n Providing modeling assistance to domain modelers is a prominent challenge in incorporating humans into planning processes. Many efforts have been devoted to this direction in classical planning, however, only few works have been done in hierarchical planning. In this thesis, we will study a methodology for providing modeling assistance in HTN planning, which is the most commonly used hierarchical planning framework. Particularly, we will address two bottleneck problems for this purpose, namely domain model validation and domain model refinements. For the former one, we propose an approach based on plan verification, and for the latter, we view it as a model reconciliation problem and will study a novel approach for solving it.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Neural Network Action Policy Verification via Predicate Abstraction.\n \n \n \n \n\n\n \n Marcel Vinzent.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 41–45, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Neural paper\n  \n \n \n \"Neural presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 6 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-10,\n  author           = {Marcel Vinzent},\n  title            = {Neural Network Action Policy Verification via Predicate Abstraction},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {41--45},\n  abstract         = {Neural networks (NN) are an increasingly important representation of action policies. With their application for realtime decision-making in safety critical areas, like, e.g., autonomous driving, it arises the need to gain trust in the applied policies. The ultimate goal to gain this trust is through formal verification of the policy-induced behavior. This is a challenging endeavor as it compounds the state space explosion with the difficulty of analyzing even single NN decision episodes. In our work, we make a contribution to cope with this challenge. We approach safety verification through (overapproximating) abstract reachability analysis. We compute predicate abstractions of the policy-restricted state space; expressing the abstract transition computation as a satisfiability modulo theories (SMT) problem, and devise a range of algorithmic enhancements to avoid costly calls to SMT. First empirical results show that our approach can outperform competing approaches. Future work will further enhance the technique and extend it to support probabilistic settings.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_352.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=ronje0VmWog&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n\n
\n
\n\n\n
\n Neural networks (NN) are an increasingly important representation of action policies. With their application for realtime decision-making in safety critical areas, like, e.g., autonomous driving, it arises the need to gain trust in the applied policies. The ultimate goal to gain this trust is through formal verification of the policy-induced behavior. This is a challenging endeavor as it compounds the state space explosion with the difficulty of analyzing even single NN decision episodes. In our work, we make a contribution to cope with this challenge. We approach safety verification through (overapproximating) abstract reachability analysis. We compute predicate abstractions of the policy-restricted state space; expressing the abstract transition computation as a satisfiability modulo theories (SMT) problem, and devise a range of algorithmic enhancements to avoid costly calls to SMT. First empirical results show that our approach can outperform competing approaches. Future work will further enhance the technique and extend it to support probabilistic settings.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Plan Recognition.\n \n \n \n \n\n\n \n Kristýna Pantůčková.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 46–49, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Plan paper\n  \n \n \n \"Plan presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 6 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-11,\n  author           = {Kristýna Pantůčková},\n  title            = {Plan Recognition},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {46--49},\n  abstract         = {The topic of the dissertation is plan recognition. Plan recognition is the task of recognizing the goal of an agent based on the observed actions. The aim of the current research is to develop an efficient approach to plan recognition in hierarchical task networks (HTN). We intend to improve the performance of existing parsing-based approach by heuristics based on landmarks.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_357.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=Uc757Bx_9tE&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n\n
\n
\n\n\n
\n The topic of the dissertation is plan recognition. Plan recognition is the task of recognizing the goal of an agent based on the observed actions. The aim of the current research is to develop an efficient approach to plan recognition in hierarchical task networks (HTN). We intend to improve the performance of existing parsing-based approach by heuristics based on landmarks.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n Probabilistic Replanning with Guarantees.\n \n \n \n \n\n\n \n Johannes Schmalz.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 50–53, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"Probabilistic paper\n  \n \n \n \"Probabilistic presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 5 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-12,\n  author           = {Johannes Schmalz},\n  title            = {Probabilistic Replanning with Guarantees},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {50--53},\n  abstract         = {State-of-the-art probabilistic replanners are solvers for probabilistic planning problems that offer a very efficient means to generate a sub-optimal solution quickly, and in an any-time fashion improve on it. Unfortunately, current approaches do so at the cost of guarantees, i.e. the solution may not be optimal, and it can not guarantee its solution will lead to the goal with certainty. To address this issue we introduce CoGNeRe, a novel probabilistic replanner that uses techniques from operations research to provide guarantees and flexibility that previous replanners can not offer.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_355.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=cmX54uOjbWM&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n\n
\n
\n\n\n
\n State-of-the-art probabilistic replanners are solvers for probabilistic planning problems that offer a very efficient means to generate a sub-optimal solution quickly, and in an any-time fashion improve on it. Unfortunately, current approaches do so at the cost of guarantees, i.e. the solution may not be optimal, and it can not guarantee its solution will lead to the goal with certainty. To address this issue we introduce CoGNeRe, a novel probabilistic replanner that uses techniques from operations research to provide guarantees and flexibility that previous replanners can not offer.\n
\n\n\n
\n\n\n
\n \n\n \n \n \n \n \n \n La VIDA: A System for Value and Identity Driven Autonomous Agent Behavior in Vir- tual World Scenarios.\n \n \n \n \n\n\n \n Ursula Addison.\n\n\n \n\n\n\n In Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022), pages 54–57, 2022. \n \n\n\n\n
\n\n\n\n \n \n \"La paper\n  \n \n \n \"La presentation\n  \n \n\n \n\n \n link\n  \n \n\n bibtex\n \n\n \n  \n \n abstract \n \n\n \n  \n \n 4 downloads\n \n \n\n \n \n \n \n \n \n \n\n  \n \n \n\n\n\n
\n 
@InProceedings{ICAPSDC2022paper-13,\n  author           = {Ursula Addison},\n  title            = {La VIDA: A System for Value and Identity Driven Autonomous Agent Behavior in Vir-\ntual World Scenarios},\n  booktitle        = {Proceedings of the 20th ICAPS Doctoral Consortium (ICAPS DC 2022)},\n  year             = {2022},\n  pages            = {54--57},\n  abstract         = {There are a great variety of systems that control agents using models of human behavior. However, often times agent action is still a reflection of the system designer’s expectations and desired outcome. But, how would agents behave if they had an identity similar to a human? What goals would be formed and how would those goals be realized as actions? We would like to produce agent behavior using these questions as guidance for our work. To this end we investigate how long-term autonomy is influenced by an agent’s identity and how these findings can be used to direct the behavior of artificial agents. For our system la VIDA, first we will create a model for human identity and ultimately integrate it with a Goal-Driven Autonomy (GDA) system at the drive level.},\n  url_paper        = {https://icaps22.icaps-conference.org/dc/ICAPS_2022_paper_354.pdf},\n  url_presentation = {https://www.youtube.com/watch?v=Ia3DkJkjAMI&list=PLj-ZdQ5rfSEqD1SztBXJppdjIE9CQQfzV}\n}\n
\n
\n\n\n
\n There are a great variety of systems that control agents using models of human behavior. However, often times agent action is still a reflection of the system designer’s expectations and desired outcome. But, how would agents behave if they had an identity similar to a human? What goals would be formed and how would those goals be realized as actions? We would like to produce agent behavior using these questions as guidance for our work. To this end we investigate how long-term autonomy is influenced by an agent’s identity and how these findings can be used to direct the behavior of artificial agents. For our system la VIDA, first we will create a model for human identity and ultimately integrate it with a Goal-Driven Autonomy (GDA) system at the drive level.\n
\n\n\n
\n\n\n\n\n\n
\n
\n\n\n\n\n
\n\n\n \n\n \n \n \n \n\n
\n"}; document.write(bibbase_data.data);