Multi-label spacecraft electrical signal classification method based on DBN and random forest
Fig 10
Classification accuracy using different algorithms before dimension reduction.
Before the dimension reduction of the data, the classification accuracy of the four algorithms is shown in Fig 10. Comparing the accuracy of the algorithm, the RF algorithm has the highest classification accuracy. Due to the limitation of the multi-class classification problems, random forest has better tolerance, excellent performance, and the highest classification accuracy before dimension reduction.