Bilinear Model
Bilinear models are computational tools that leverage the interaction between pairs of feature vectors to improve performance in various machine learning tasks. Current research focuses on developing efficient bilinear architectures, such as lightweight convolutional networks and bilinear multilayer perceptrons, and applying them to diverse domains including image classification, natural language processing, and knowledge graph completion. These models offer advantages in terms of accuracy, interpretability, and computational efficiency, leading to improved performance in applications ranging from medical image analysis to mathematical proof verification and control systems.
Papers
October 24, 2024
September 17, 2024
June 6, 2024
May 10, 2024
September 25, 2023
February 18, 2023
September 20, 2022
March 16, 2022