Zhang, C., Geng, Y., Han, Z., Liu, Y., Fu, H., & Hu, Q. (2022). Autoencoder in Autoencoder Networks. IEEE Transactions on Neural Networks and Learning Systems, 1–13. https://doi.org/10.1109/tnnls.2022.3189239
Modeling complex correlations on multiview data is still challenging, especially for high-dimensional features with possible noise. To address this issue, we propose a novel unsupervised multiview representation learning (UMRL) algorithm, termed autoencoder in autoencoder networks (AE 2 -Nets). The proposed framework effectively encodes information from high-dimensional heterogeneous data into a compact and informative representation with the proposed bidirectional encoding strategy. Speciﬁcally, the proposed AE 2 -Nets conduct encoding in two directions: the inner-AE-networks extract view-speciﬁc intrinsic information (forward encoding), while the outer-AE-networks integrate this view-speciﬁc intrinsic information from different views into a latent representation (backward encoding). For the nested architecture, we further provide a probabilistic explanation and extension from hierarchical variational autoencoder. The forward–backward strategy ﬂexibly addresses high-dimensional (noisy) features within each view and encodes complementarity across multiple views in a uniﬁed framework. Extensive results on benchmark datasets validate the advantages compared to the state-of-the-art algorithms.
This research / project is supported by the AISG - AISG Tech Challenge Funding
Grant Reference no. : AISG2-TC-2021-003
This work was supported in part by the National Key Research and Development Program of China under Grant 2019YFB2101900; in part by the National Natural Science Foundation of China under Grant 61976151, Grant 61925602, and Grant 61732011; in part by the Natural Science Foundation of Tianjin under Grant 19JCYBJC15200