Random forest-compressed sensing method for pulsar recognition

For the pulsar recognition, there are several problems such as large data volume, high feature dimension, large computational complexity and low recognition rate. A random forest-compressed sensing method for pulsar recognition is developed. Compressed sensing can achieve the fast compression of the raw pulsar data. And random forest has a good classification capability. In this paper, the random forest algorithm is used as the reconstruction algorithm in compressed sensing. Based on their advantages, a random forest-compressed sensing algorithm is developed to apply in the pulsar recognition. The experimental results show that compared with the traditional pulsar recognition algorithms, the proposed algorithm has a higher recognition rate and a smaller computational complexity.