{"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/discovering-influential-factors-in","title":"Discovering Influential Factors in Variational Autoencoder","arxiv_id":"1809.01804","date":"2018-09-06","proceeding":null,"authors":["Shiqi Liu","Jingxin Liu","Qian Zhao","Xiangyong Cao","Huibin Li","Hongy-ing Meng","Sheng Liu","Deyu Meng"],"abstract":"In the field of machine learning, it is still a critical issue to identify\nand supervise the learned representation without manually intervening or\nintuition assistance to extract useful knowledge or serve for the downstream\ntasks. In this work, we focus on supervising the influential factors extracted\nby the variational autoencoder(VAE). The VAE is proposed to learn independent\nlow dimension representation while facing the problem that sometimes pre-set\nfactors are ignored. We argue that the mutual information of the input and each\nlearned factor of the representation plays a necessary indicator of discovering\nthe influential factors. We find the VAE objective inclines to induce mutual\ninformation sparsity in factor dimension over the data intrinsic dimension and\nresults in some non-influential factors whose function on data reconstruction\ncould be ignored. We show mutual information also influences the lower bound of\nVAE's reconstruction error and downstream classification task. To make such\nindicator applicable, we design an algorithm for calculating the mutual\ninformation for VAE and prove its consistency. Experimental results on MNIST,\nCelebA and DEAP datasets show that mutual information can help determine\ninfluential factors, of which some are interpretable and can be used to further\ngeneration and classification tasks, and help discover the variant that\nconnects with emotion on DEAP dataset.","url_abs":"http://arxiv.org/abs/1809.01804v2","url_pdf":"http://arxiv.org/pdf/1809.01804v2.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":"discovering-influential-factors-in","repo_url":"https://github.com/647LiuSQ/Discovering-influential-factors-in-variational-autoencoders","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"classification","task_name":"General Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}