Abstract
This paper proposes a new open-ended learning framework which aims at implementing an autonomous agent using intrinsic motivations (IM) at two different levels.
At the first level, the IM paradigm is exploited by the agent to learn new operational skills, described in terms of sub-symbolic options. After discovering the options, the agent iteratively: (1) executes them to explore the world, collecting the necessary data and (2) automatically abstracts the collected data into a high-level representation of the domain, expressed in PPDDL language.
At the second level, the IM paradigm is used to exploit the abstracted representation of the domain by identifying particular symbolic states deemed promising according to a specific criterium, which in the present work is the farthest distance covered by the agent (i.e., the most promising states are those that rest at the frontier of the visited space). Once these states are identified, they can be successively reached through an internally generated high-level plan and used as promising starting points for discovering new knowledge.
The presented framework is tested in the so-called Treasure Game domain described in the recent literature. The tests we have performed show that the proposed idea of implementing intrinsic motivations at two different levels of abstraction facilitates the discovery of new knowledge, compared to a previous approach proposed in the literature.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
The mask is the list of state variables changed by a specific option [16].
References
Baldassarre, G., Mirolli, M.: Intrinsically Motivated Learning in Natural and Artificial Systems. Springer, Heidelberg (2013). https://doi.org/10.1007/978-3-642-32375-1
Barto, A.G., Mahadevan, S.: Recent advances in hierarchical reinforcement learning. Disc. Event Dyn. Syst. 13(1), 41–77 (2003)
Bellemare, M., Srinivasan, S., Ostrovski, G., Schaul, T., Saxton, D., Munos, R.: Unifying count-based exploration and intrinsic motivation. Adv. Neural Inf. Process. Syst. 29 (2016)
Bengio, Y., Louradour, J., Collobert, R., Weston, J.: Curriculum learning. In: Proceedings of the 26th Annual International Conference on Machine Learning, pp. 41–48 (2009)
Blaes, S., Vlastelica Pogančić, M., Zhu, J., Martius, G.: Control what you can: intrinsically motivated task-planning agent. Adv. Neural Inf. Process. Syst. 32 (2019)
Bonet, B., Geffner, H.: MGPT: a probabilistic planner based on heuristic search. J. Artif. Int. Res. 24(1), 933–944 (2005)
Campari, T., Lamanna, L., Traverso, P., Serafini, L., Ballan, L.: Online learning of reusable abstract models for object goal navigation. In: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 14850–14859 (2022). https://doi.org/10.1109/CVPR52688.2022.01445
Colas, C., Fournier, P., Chetouani, M., Sigaud, O., Oudeyer, P.Y.: Curious: intrinsically motivated modular multi-goal reinforcement learning. In: International Conference on Machine Learning, pp. 1331–1340. PMLR (2019)
Cortes, C., Vapnik, V.: Support-vector networks. Mach. Learn. 20(3), 273–297 (1995)
Ester, M., Kriegel, H.P., Sander, J., Xu, X.: A density-based algorithm for discovering clusters in large spatial databases with noise. In: Proceedings of the Second International Conference on Knowledge Discovery and Data Mining, KDD 1996, pp. 226–231. AAAI Press (1996)
Forestier, S., Portelas, R., Mollard, Y., Oudeyer, P.Y.: Intrinsically motivated goal exploration processes with automatic curriculum learning. arXiv preprint arXiv:1708.02190 (2017)
Frank, M., Leitner, J., Stollenga, M., Förster, A., Schmidhuber, J.: Curiosity driven reinforcement learning for motion planning on humanoids. Front. Neurorobot. 7, 25 (2014)
Ghallab, M., et al.: PDDL–the planning domain definition language (1998). http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.37.212
Jong, N.K., Hester, T., Stone, P.: The utility of temporal abstraction in reinforcement learning. In: AAMAS, no. 1, pp. 299–306. Citeseer (2008)
Konidaris, G., Barto, A.G.: Skill discovery in continuous reinforcement learning domains using skill chaining. Adv. Neural Inf. Process. Syst., 1015–1023 (2009)
Konidaris, G., Kaelbling, L.P., Lozano-Perez, T.: From skills to symbols: learning symbolic representations for abstract high-level planning. J. Artif. Intell. Res. 61, 215–289 (2018). http://lis.csail.mit.edu/pubs/konidaris-jair18.pdf
Lamanna, L., et al.: Planning for learning object properties. In: Williams, B., Chen, Y., Neville, J. (eds.) Thirty-Seventh AAAI Conference on Artificial Intelligence, AAAI 2023, Thirty-Fifth Conference on Innovative Applications of Artificial Intelligence, IAAI 2023, Thirteenth Symposium on Educational Advances in Artificial Intelligence, EAAI 2023, Washington, DC, USA, 7–14 February 2023, pp. 12005–12013. AAAI Press (2023). http://ojs.aaai.org/index.php/AAAI/article/view/26416
Lamanna, L., Serafini, L., Saetti, A., Gerevini, A., Traverso, P.: Online grounding of symbolic planning domains in unknown environments. In: Kern-Isberner, G., Lakemeyer, G., Meyer, T. (eds.) Proceedings of the 19th International Conference on Principles of Knowledge Representation and Reasoning, KR 2022, Haifa, Israel, 31 July–5 August 2022 (2022). http://proceedings.kr.org/2022/53/
Machado, M.C., Bellemare, M.G., Bowling, M.: A laplacian framework for option discovery in reinforcement learning. arXiv preprint arXiv:1703.00956 (2017)
Mann, T.A., Mannor, S., Precup, D.: Approximate value iteration with temporally extended actions. J. Artif. Intell. Res. 53, 375–438 (2015)
Nau, D., Ghallab, M., Traverso, P.: Automated Planning: Theory & Practice. Morgan Kaufmann Publishers Inc., San Francisco (2004)
Niel, R., Wiering, M.A.: Hierarchical reinforcement learning for playing a dynamic dungeon crawler game. In: 2018 IEEE Symposium Series on Computational Intelligence (SSCI), pp. 1159–1166. IEEE (2018)
Oddi, A., et al.: Integrating open-ended learning in the sense-plan-act robot control paradigm. In: ECAI 2020, the 24th European Conference on Artificial Intelligence (2020)
Oudeyer, P.Y., Kaplan, F., Hafner, V.: Intrinsic motivation systems for autonomous mental development. IEEE Trans. Evol. Comput. 11(2), 265–286 (2007)
Parisi, S., Dean, V., Pathak, D., Gupta, A.: Interesting object, curious agent: learning task-agnostic exploration. Adv. Neural. Inf. Process. Syst. 34, 20516–20530 (2021)
Parzen, E.: On estimation of a probability density function and mode. Ann. Math. Stat. 33(3), 1065–1076 (1962)
Pedregosa, F., et al.: Scikit-learn: machine learning in python. J. Mach. Learn. Res. 12(null), 2825–2830 (2011)
Romero, A., Baldassarre, G., Duro, R.J., Santucci, V.G.: Analysing autonomous open-ended learning of skills with different interdependent subgoals in robots. In: 2021 20th International Conference on Advanced Robotics (ICAR), pp. 646–651. IEEE (2021)
Romero, A., Baldassarre, G., Duro, R.J., Santucci, V.G.: Autonomous learning of multiple curricula with non-stationary interdependencies. In: 2022 IEEE International Conference on Development and Learning (ICDL), pp. 272–279. IEEE (2022)
Rosenblatt, M.: Remarks on some nonparametric estimates of a density function. Ann. Math. Stat. 27(3), 832–837 (1956)
Sanner, S.: Relational dynamic influence diagram language (rddl): language description (2010). http://users.cecs.anu.edu.au/~ssanner/IPPC_2011/RDDL.pdf
Santucci, V.G., Baldassarre, G., Mirolli, M.: Grail: a goal-discovering robotic architecture for intrinsically-motivated learning. IEEE Trans. Cogn. Dev. Syst. 8(3), 214–231 (2016)
Santucci, V.G., Oudeyer, P.Y., Barto, A., Baldassarre, G.: Intrinsically motivated open-ended learning in autonomous robots. Front. Neurorobot. 13, 115 (2020)
Sartor, G., Zollo, D., Mayer, M.C., Oddi, A., Rasconi, R., Santucci, V.G.: Autonomous generation of symbolic knowledge via option discovery. In: Proceedings of the 9th Italian workshop on Planning and Scheduling (IPS 2021), vol. 3065. CEUR Workshop Proceedings. CEUR-WS.org (2021)
Seepanomwan, K., Santucci, V.G., Baldassarre, G.: Intrinsically motivated discovered outcomes boost user’s goals achievement in a humanoid robot. In: 2017 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob), pp. 178–183 (2017)
Singh, S., Barto, A.G., Chentanez, N.: Intrinsically motivated reinforcement learning. In: Proceedings of the 17th International Conference on Neural Information Processing Systems, NIPS 2004, pp. 1281–1288. MIT Press, Cambridge (2004)
Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT press, Cambridge (2018)
Sutton, R.S., Precup, D., Singh, S.: Between MDPS and semi-MDPS: a framework for temporal abstraction in reinforcement learning. Artif. Intell. 112(1), 181–211 (1999)
Younes, H., Littman, M.: PPDDL1.0: An Extension to PDDL for Expressiong Planning Domains with Probabilistic Effects. Technical report, Carnegie Mellon University, CMU-CS-04-167 (2004)
Acknowledgements
This work has been supported by the European Union’s Horizon 2020, research and innovation programme under GA 101070381 (‘PILLAR-Robots - Purposeful Intrinsically motivated Lifelong Learning Autonomous Robots’) and PNRR MUR project PE0000013-FAIR.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Sartor, G., Oddi, A., Rasconi, R., Santucci, V.G. (2023). Intrinsically Motivated High-Level Planning for Agent Exploration. In: Basili, R., Lembo, D., Limongelli, C., Orlandini, A. (eds) AIxIA 2023 – Advances in Artificial Intelligence. AIxIA 2023. Lecture Notes in Computer Science(), vol 14318. Springer, Cham. https://doi.org/10.1007/978-3-031-47546-7_9
Download citation
DOI: https://doi.org/10.1007/978-3-031-47546-7_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-47545-0
Online ISBN: 978-3-031-47546-7
eBook Packages: Computer ScienceComputer Science (R0)