2018/09/07 by Tengfei Ma, Jie Chen, Ma, Tengfei +3 · 3 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · #Bioinformatics and Genomic Networks #Biomedical Text Mining and Ontologies #Data Visualization and Analytics #FOS: Computer and information sciences #Gene expression and cancer classification #Machine Learning (cs.LG) #Machine Learning (stat.ML)
paper · pdf · doi:10.48550/arxiv.1809.02630
openalex publication_date 2018/09/07 · openalex created_date 2022/08/03 · openalex updated_date 2026/07/28
Deep generative models have achieved remarkable success in various data\ndomains, including images, time series, and natural languages. There remain,\nhowever, substantial challenges for combinatorial structures, including graphs.\nOne of the key challenges lies in the difficulty of ensuring semantic validity\nin context. For examples, in molecular graphs, the number of bonding-electron\npairs must not exceed the valence of an atom; whereas in protein interaction\nnetworks, two proteins may be connected only when they belong to the same or\ncorrelated gene ontology terms. These constraints are not easy to be\nincorporated into a generative model. In this work, we propose a regularization\nframework for variational autoencoders as a step toward semantic validity. We\nfocus on the matrix representation of graphs and formulate penalty terms that\nregularize the output distribution of the decoder to encourage the satisfaction\nof validity constraints. Experimental results confirm a much higher likelihood\nof sampling valid graphs in our approach, compared with others reported in the\nliterature.\n