Fast Computation of Sample Entropy Based on the Iteration of Neighborhood Matrix

In order to reduce the computational complexity of sample entropy, a fast method is presented based on the proposed neighborhood matrix. The neighborhood matrix of highdimensional signal vector can be efficiently deduced from that of the lower dimensional signal vector. The efficiency of the fast algorithm is analyzed theoretically, and its validity is proved by experiments. Compared with the traditional algorithm, the computation time is significantly reduced. The experimental results show that the proposed method is effective, which facilitates real-time applications of the sample entropy.