%0 Conference Paper %F Oral %T A new neural network feature importance method: Application to mobile robots controllers gain tuning %+ Laboratoire de Robotique Interactive (LRI) %+ Technologies et systèmes d'information pour les agrosystèmes (UR TSCF) %A Hill, Ashley %A Lucet, Eric %A Lenain, Roland %< avec comité de lecture %B ICINCO 2020, 17th International Conference on Informatics in Control, Automation and Robotics %C Paris, France %8 2020-07-07 %D 2020 %K Machine Learning %K Neural Network %K Robotics %K Mobile Robot %K Control Theory %K Gain Tuning %K Adaptive Control %K Explainable Artificial Intelligence %Z Computer Science [cs]/Artificial Intelligence [cs.AI] %Z Engineering Sciences [physics]/AutomaticConference papers %X This paper proposes a new approach for feature importance of neural networks and subsequently a methodology to determine useful sensor information in high performance controllers, using a trained neural network that predicts the quasi-optimal gain in real time. The neural network is trained using the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) algorithm, in order to lower a given objective function. The important sensor information for robotic control are determined using the described methodology. Then a proposed improvement to the tested control law is given, and compared with the neural network's gain prediction method for real time gain tuning. As a results, crucial information about the importance of a given sensory information for robotic control is determined, and shown to improve the performance of existing controllers. %G English %2 https://cea.hal.science/cea-03314585/document %2 https://cea.hal.science/cea-03314585/file/2020_ICINCO_Hill.pdf %L cea-03314585 %U https://cea.hal.science/cea-03314585 %~ CEA %~ DRT %~ CEA-UPSAY %~ TDS-MACS %~ UNIV-PARIS-SACLAY %~ LIST %~ INRAE %~ UNIVERSITE-PARIS-SACLAY %~ GS-ENGINEERING %~ GS-COMPUTER-SCIENCE %~ TSCF %~ MATHNUM