TJU Scholars Advance Research on Multi-Modal Machine Learning

Recently, the paper titled Deep Partial Multi-View Learning from the lab of machine learning and data mining at the College of Intelligence and Computing was accepted byIEEE Transaction on Pattern Analysis and Machine Intelligence(SCI, IF: 17.861), a top international journal in Artificial Intelligence. Important progress has been made in the methods and theories for integration on data with multiple modalities.

Although multi-view learning has made significant progress over the past few decades, it is still challenging due to the difficulty in modeling complex correlations among different views, especially under the context of view missing. The work proposes a novel framework to fully and flexibly take advantage of multiple partial views. This is the first formal definition of completeness and versatility for multi-view representation and theoretical results for the versatility of latent representations.

By the College of Intelligence and Computing

Editor: Eva Yin