Matches in SemOpenAlex for { <https://semopenalex.org/work/W3087163951> ?p ?o ?g. }
- W3087163951 abstract "Black-box artificial intelligence (AI) induction methods such as deep reinforcement learning (DRL) are increasingly being used to find optimal policies for a given control task. Although policies represented using a black-box AI are capable of efficiently executing the underlying control task and achieving optimal closed-loop performance -- controlling the agent from initial time step until the successful termination of an episode, the developed control rules are often complex and neither interpretable nor explainable. In this paper, we use a recently proposed nonlinear decision-tree (NLDT) approach to find a hierarchical set of control rules in an attempt to maximize the open-loop performance for approximating and explaining the pre-trained black-box DRL (oracle) agent using the labelled state-action dataset. Recent advances in nonlinear optimization approaches using evolutionary computation facilitates finding a hierarchical set of nonlinear control rules as a function of state variables using a computationally fast bilevel optimization procedure at each node of the proposed NLDT. Additionally, we propose a re-optimization procedure for enhancing closed-loop performance of an already derived NLDT. We evaluate our proposed methodologies on four different control problems having two to four discrete actions. In all these problems our proposed approach is able to find simple and interpretable rules involving one to four non-linear terms per rule, while simultaneously achieving on par closed-loop performance when compared to a trained black-box DRL agent. The obtained results are inspiring as they suggest the replacement of complicated black-box DRL policies involving thousands of parameters (making them non-interpretable) with simple interpretable policies. Results are encouraging and motivating to pursue further applications of proposed approach in solving more complex control tasks." @default.
- W3087163951 created "2020-09-25" @default.
- W3087163951 creator A5015550681 @default.
- W3087163951 creator A5061604841 @default.
- W3087163951 creator A5068379698 @default.
- W3087163951 creator A5088394271 @default.
- W3087163951 creator A5088648447 @default.
- W3087163951 date "2020-09-20" @default.
- W3087163951 modified "2023-09-27" @default.
- W3087163951 title "Interpretable-AI Policies using Evolutionary Nonlinear Decision Trees for Discrete Action Systems." @default.
- W3087163951 cites W11104910 @default.
- W3087163951 cites W1499669280 @default.
- W3087163951 cites W1568834902 @default.
- W3087163951 cites W1585546346 @default.
- W3087163951 cites W1595498733 @default.
- W3087163951 cites W2000295051 @default.
- W3087163951 cites W2103862045 @default.
- W3087163951 cites W2120346334 @default.
- W3087163951 cites W2124175081 @default.
- W3087163951 cites W2165558283 @default.
- W3087163951 cites W2173248099 @default.
- W3087163951 cites W225560312 @default.
- W3087163951 cites W2543580944 @default.
- W3087163951 cites W2554751554 @default.
- W3087163951 cites W2614367549 @default.
- W3087163951 cites W2728455196 @default.
- W3087163951 cites W2729084506 @default.
- W3087163951 cites W2736601468 @default.
- W3087163951 cites W2796284132 @default.
- W3087163951 cites W2883535494 @default.
- W3087163951 cites W2892364115 @default.
- W3087163951 cites W2949608212 @default.
- W3087163951 cites W2952523895 @default.
- W3087163951 cites W2962957031 @default.
- W3087163951 cites W2963672746 @default.
- W3087163951 cites W2964043796 @default.
- W3087163951 cites W3008495039 @default.
- W3087163951 cites W3021613070 @default.
- W3087163951 cites W3037467471 @default.
- W3087163951 cites W3046797186 @default.
- W3087163951 cites W3139377883 @default.
- W3087163951 hasPublicationYear "2020" @default.
- W3087163951 type Work @default.
- W3087163951 sameAs 3087163951 @default.
- W3087163951 citedByCount "1" @default.
- W3087163951 countsByYear W30871639512020 @default.
- W3087163951 crossrefType "posted-content" @default.
- W3087163951 hasAuthorship W3087163951A5015550681 @default.
- W3087163951 hasAuthorship W3087163951A5061604841 @default.
- W3087163951 hasAuthorship W3087163951A5068379698 @default.
- W3087163951 hasAuthorship W3087163951A5088394271 @default.
- W3087163951 hasAuthorship W3087163951A5088648447 @default.
- W3087163951 hasConcept C113174947 @default.
- W3087163951 hasConcept C11413529 @default.
- W3087163951 hasConcept C115903868 @default.
- W3087163951 hasConcept C119857082 @default.
- W3087163951 hasConcept C121332964 @default.
- W3087163951 hasConcept C126255220 @default.
- W3087163951 hasConcept C127413603 @default.
- W3087163951 hasConcept C134306372 @default.
- W3087163951 hasConcept C137836250 @default.
- W3087163951 hasConcept C154945302 @default.
- W3087163951 hasConcept C158622935 @default.
- W3087163951 hasConcept C177264268 @default.
- W3087163951 hasConcept C199360897 @default.
- W3087163951 hasConcept C201995342 @default.
- W3087163951 hasConcept C2780451532 @default.
- W3087163951 hasConcept C3309286 @default.
- W3087163951 hasConcept C33923547 @default.
- W3087163951 hasConcept C41008148 @default.
- W3087163951 hasConcept C55166926 @default.
- W3087163951 hasConcept C62520636 @default.
- W3087163951 hasConcept C84525736 @default.
- W3087163951 hasConcept C94966114 @default.
- W3087163951 hasConcept C97541855 @default.
- W3087163951 hasConceptScore W3087163951C113174947 @default.
- W3087163951 hasConceptScore W3087163951C11413529 @default.
- W3087163951 hasConceptScore W3087163951C115903868 @default.
- W3087163951 hasConceptScore W3087163951C119857082 @default.
- W3087163951 hasConceptScore W3087163951C121332964 @default.
- W3087163951 hasConceptScore W3087163951C126255220 @default.
- W3087163951 hasConceptScore W3087163951C127413603 @default.
- W3087163951 hasConceptScore W3087163951C134306372 @default.
- W3087163951 hasConceptScore W3087163951C137836250 @default.
- W3087163951 hasConceptScore W3087163951C154945302 @default.
- W3087163951 hasConceptScore W3087163951C158622935 @default.
- W3087163951 hasConceptScore W3087163951C177264268 @default.
- W3087163951 hasConceptScore W3087163951C199360897 @default.
- W3087163951 hasConceptScore W3087163951C201995342 @default.
- W3087163951 hasConceptScore W3087163951C2780451532 @default.
- W3087163951 hasConceptScore W3087163951C3309286 @default.
- W3087163951 hasConceptScore W3087163951C33923547 @default.
- W3087163951 hasConceptScore W3087163951C41008148 @default.
- W3087163951 hasConceptScore W3087163951C55166926 @default.
- W3087163951 hasConceptScore W3087163951C62520636 @default.
- W3087163951 hasConceptScore W3087163951C84525736 @default.
- W3087163951 hasConceptScore W3087163951C94966114 @default.
- W3087163951 hasConceptScore W3087163951C97541855 @default.
- W3087163951 hasLocation W30871639511 @default.
- W3087163951 hasOpenAccess W3087163951 @default.