Research on the Evaluation Model of Dance Movement Recognition and Automatic Generation Based on Long Short-Term Memory

Table 2

Three characteristics of self-encoder.

Characteristic	Content
Data correlation	Refers that the self-encoder can only compress data similar to its previous training data.
Data lossiness [23]	Compared with the original input, the output of self-encoder will lose information when decompressing, so self-encoder is a data lossy compression algorithm.
Automatic learning [24]	Automatic encoders learn automatically from data samples, which mean that it is easy to train a specific encoder for the input of a specified class without completing any new work.