An Efficient and Effective Second-Order Training Algorithm for LSTM-based Adaptive Learning
This addresses efficient online learning for LSTM networks, offering practical speed and accuracy gains, though it appears incremental as an enhancement to existing EKF methods.
The paper tackles adaptive nonlinear regression with LSTM networks by introducing an Extended Kalman filter-based second-order training algorithm, demonstrating 10-45% accuracy improvements over Adam, RMSprop, and DEKF, and comparable performance to EKF with 10-15 times faster run-time.
We study adaptive (or online) nonlinear regression with Long-Short-Term-Memory (LSTM) based networks, i.e., LSTM-based adaptive learning. In this context, we introduce an efficient Extended Kalman filter (EKF) based second-order training algorithm. Our algorithm is truly online, i.e., it does not assume any underlying data generating process and future information, except that the target sequence is bounded. Through an extensive set of experiments, we demonstrate significant performance gains achieved by our algorithm with respect to the state-of-the-art methods. Here, we mainly show that our algorithm consistently provides 10 to 45\% improvement in the accuracy compared to the widely-used adaptive methods Adam, RMSprop, and DEKF, and comparable performance to EKF with a 10 to 15 times reduction in the run-time.