Policy Search in a Space of Simple Closed-form Formulas: Towards Interpretability of Reinforcement Learning

By Francis Maes, Raphael Fonteneau, Louis Wehenkel, Damien Ernst. 2012

30 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Management Science and Operations Research

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. Chao Yu
2. Jiming Liu
3. Shamim Nemati
4. Steffen Udluft
5. Thomas A. Runkler
6. Yashesh D. Dhebar
7. Finale Doshi-Velez
8. David A. Sontag

In Management Science and Operations Research

1. Finale Doshi-Velez
2. Leo Anthony Celi
3. Shamim Nemati
4. Fredrik D. Johansson
5. Kalyanmoy Deb
6. Marek Petrik
7. Ronny Luss
8. Steffen Udluft

Citations in the corpus

• Evaluating Reinforcement Learning Algorithms in Observational Health Settings (2018, background): “More recently some works have also specifi-cally attempted to address the problem of interpretability in RL algorithms [Maes et al., 2012, Hein et al., 2017].”
• Toward Interpretable-AI Policies Using Evolutionary Nonlinear Decision Trees for Discrete-Action Systems (2020, methodology): “Maes et al. [16] constructed a set of interpretable index-based policies and used a multi-arm bandit procedure to select a high performing index-based policy.”
• Reinforcement Learning in Healthcare: A Survey (2019, methodology): “he problem of interpretability in RL algorithms. There are a variety of ways to realize interpretability of learned policy, by either using small, closed-form formulas to compute index-based policies [333], using program synthesis to learn higher-level symbolic interpretable representations of learn”
• Particle swarm optimization for generating interpretable fuzzy reinforcement learning policies (2016, background): “On the other hand, the search for interpretable RL policies is of significant academic and industrial interest [22].”
• Interpretable Policies for Reinforcement Learning by Genetic Programming (2017, background): “Despite the search of interpretable RL policies being of high academic and industrial interest, little has been published concerning human interpretable and understandable policies trained by data driven learning methods (Maes, Fonteneau, Wehenkel, and Ernst, 2012).”
• Interpretable Policies for Dynamic Product Recommendations (2016, background): “Recent work on interpretable policies in reinforcement learning [17] proposes to use policy search methods.”