Gradient-Based Neuroplastic Adaptation for Concurrent Optimization of Neuro-Fuzzy Networks
This work addresses the problem of inefficient and brittle NFN design for researchers and practitioners, representing an incremental improvement by optimizing parameters and structure simultaneously.
The paper tackles the challenge of systematically designing neuro-fuzzy networks (NFNs) by proposing a gradient-based neuroplastic adaptation method for concurrent optimization of parameters and structure, enabling NFNs to be trained via online reinforcement learning to proficiently play challenging scenarios in the vision-based video game DOOM.
Neuro-fuzzy networks (NFNs) are transparent, symbolic, and universal function approximations that perform as well as conventional neural architectures, but their knowledge is expressed as linguistic IF-THEN rules. Despite these advantages, their systematic design process remains a challenge. Existing work will often sequentially build NFNs by inefficiently isolating parametric and structural identification, leading to a premature commitment to brittle and subpar architecture. We propose a novel application-independent approach called gradient-based neuroplastic adaptation for the concurrent optimization of NFNs' parameters and structure. By recognizing that NFNs' parameters and structure should be optimized simultaneously as they are deeply conjoined, settings previously unapproachable for NFNs are now accessible, such as the online reinforcement learning of NFNs for vision-based tasks. The effectiveness of concurrently optimizing NFNs is empirically shown as it is trained by online reinforcement learning to proficiently play challenging scenarios from a vision-based video game called DOOM.