The simulated results show that the improved algorithm can greatly increase accuracy on function approximation, and particularly suitable for the training of batched data.

• 仿真结果表明,改进算法能够大大提高函数逼近的精确性,并且可用于分批数据的训练。