SCAT: Second Chance Autoencoder for Textual Data
This work addresses feature extraction in textual data for applications such as classification and topic modeling, but appears incremental as it builds on competitive learning approaches.
The authors tackled the problem of learning representative features for textual data by proposing SCAT, a k-competitive learning approach for autoencoders, which achieved outstanding performance in classification, topic modeling, and document visualization compared to existing methods like LDA and KATE.
We present a k-competitive learning approach for textual autoencoders named Second Chance Autoencoder (SCAT). SCAT selects the $k$ largest and smallest positive activations as the winner neurons, which gain the activation values of the loser neurons during the learning process, and thus focus on retrieving well-representative features for topics. Our experiments show that SCAT achieves outstanding performance in classification, topic modeling, and document visualization compared to LDA, K-Sparse, NVCTM, and KATE.