{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/from-neural-pca-to-deep-unsupervised-learning","title":"From neural PCA to deep unsupervised learning","arxiv_id":"1411.7783","date":"2014-11-28","proceeding":null,"authors":["Harri Valpola"],"abstract":"A network supporting deep unsupervised learning is presented. The network is\nan autoencoder with lateral shortcut connections from the encoder to decoder at\neach level of the hierarchy. The lateral shortcut connections allow the higher\nlevels of the hierarchy to focus on abstract invariant features. While standard\nautoencoders are analogous to latent variable models with a single layer of\nstochastic variables, the proposed network is analogous to hierarchical latent\nvariables models. Learning combines denoising autoencoder and denoising sources\nseparation frameworks. Each layer of the network contributes to the cost\nfunction a term which measures the distance of the representations produced by\nthe encoder and the decoder. Since training signals originate from all levels\nof the network, all layers can learn efficiently even in deep networks. The\nspeedup offered by cost terms from higher levels of the hierarchy and the\nability to learn invariant features are demonstrated in experiments.","url_abs":"http://arxiv.org/abs/1411.7783v2","url_pdf":"http://arxiv.org/pdf/1411.7783v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"from-neural-pca-to-deep-unsupervised-learning","repo_url":"https://github.com/AbhinavS99/Ladder-Networks-for-Sign-Languages","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"denoising","task_name":"Denoising"}],"methods":[{"method_slug":"denoising-autoencoder","method_name":"Denoising Autoencoder"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}