The strategic student approach for life-long exploration and learning

By Manuel Lopes, Pierre-Yves Oudeyer. 2012

58 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications, Management Science and Operations Research, Reinforcement Learning

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

In Computer Networks and Communications

In Management Science and Operations Research

In Reinforcement Learning

Citations in the corpus

What can the brain teach us about building artificial intelligence? (2016, background): “es between models of intrinsic motivation in developmental robotics and Deep RL systems, e.g. curiosity-driven selection of parameterized problems (Baranes and Oudeyer, 2013), of learning strategies (Lopes and Oudeyer, 2012), and combinations between intrinsic motivation and social learning (Nguyen ”
Automated Curriculum Learning for Neural Networks (2017, background): “We take inspiration from a previous work that modelled an adaptive student with a multi-armed bandit in the context of developmental learning (Lopes and Oudeyer, 2012).”
Automatic Curriculum Learning For Deep RL: A Short Survey (2020, background): “Maximizing LP values was shown optimal under the assumption of concave learning profiles [Lopes and Oudeyer, 2012].”
Teacher–Student Curriculum Learning (2017, background): “y on manually designed curricula and do not attempt to produce it automatically. The idea of using learning progress as the reward was proposed in (Oudeyer and Kaplan, 2007) and later implemented in (Lopes and Oudeyer, 2012), but their work evaluates it only on synthetic tasks. Using learning progre”
Multi-Armed Bandits for Intelligent Tutoring Systems (2013, methodology): “Each simultaneous MAB uses a bandit algorithm, derived from EXP4, presented in [39].”
GEP-PG: Decoupling Exploration and Exploitation in Deep Reinforcement Learning Algorithms (2017, methodology): “A next stage could be to extend GEP - PG towards an adaptive transition or mixing mechanism, where a multi-armed bandit could be used to dynamically switch between both learning strategies using measures like learning progress (Lopes & Oudeyer, 2012).”
Active choice of teachers, learning strategies and goals for a socially guided intrinsic motivation learner (2012, background): “Only [27] studies the framework of these questions, and only examined a toy example with discrete and finite number of states, outcomes and strategies.”
Intrinsic motivation, curiosity, and learning: Theory and applications in educational technologies. (2016, background): “Indeed, Lopes and Oudeyer (2012) showed in a theoretical study that when faced with the problem of strategically choosing which topic/exercise type to work on, selecting topics/exercises that maximize learning progress is quasi-optimal for important classes of learner models.”
The effects of task difficulty, novelty and the size of the search space on intrinsically motivated exploration (2014, background): “ORIGINAL RESEARCH ARTICLE published: 14 October 2014 doi: 10.3389/fnins.2014.00317 The effects of task difficulty, novelty and the size of the search space on intrinsically motivated exploration”