EARL: Energy-Aware Optimization of Liquid State Machines for Pervasive AI
This work addresses the problem of deploying energy-efficient on-device AI systems for pervasive applications, representing an incremental improvement through a novel hybrid optimization method.
The paper tackled the challenge of optimizing Liquid State Machines for low-power temporal processing in pervasive AI by introducing EARL, an energy-aware reinforcement learning framework that jointly optimizes accuracy and energy consumption, achieving 6-15% higher accuracy, 60-80% lower energy consumption, and up to 10x faster optimization time compared to existing methods.
Pervasive AI increasingly depends on on-device learning systems that deliver low-latency and energy-efficient computation under strict resource constraints. Liquid State Machines (LSMs) offer a promising approach for low-power temporal processing in pervasive and neuromorphic systems, but their deployment remains challenging due to high hyperparameter sensitivity and the computational cost of traditional optimization methods that ignore energy constraints. This work presents EARL, an energy-aware reinforcement learning framework that integrates Bayesian optimization with an adaptive reinforcement learning based selection policy to jointly optimize accuracy and energy consumption. EARL employs surrogate modeling for global exploration, reinforcement learning for dynamic candidate prioritization, and an early termination mechanism to eliminate redundant evaluations, substantially reducing computational overhead. Experiments on three benchmark datasets demonstrate that EARL achieves 6 to 15 percent higher accuracy, 60 to 80 percent lower energy consumption, and up to an order of magnitude reduction in optimization time compared to leading hyperparameter tuning frameworks. These results highlight the effectiveness of energy-aware adaptive search in improving the efficiency and scalability of LSMs for resource-constrained on-device AI applications.