Accelerating parallel tangent learning for neural networks through dynamic self-adaptation | IEEE Conference Publication | IEEE Xplore