Target Distribution
Target distribution research focuses on efficiently approximating and sampling from complex probability distributions, crucial for various applications like drug discovery and astronomical simulations. Current efforts center on developing novel generative models, including normalizing flows and diffusion models, often coupled with optimization techniques like annealed importance sampling and gradient flows, to overcome challenges such as representation bias and high dimensionality. These advancements improve the accuracy and efficiency of generating samples from target distributions, impacting fields ranging from machine learning and statistical physics to robotics and healthcare through improved model training and data analysis.
Papers
May 20, 2023
April 28, 2023
February 23, 2023
February 16, 2023
February 1, 2023
January 18, 2023
October 31, 2022
June 6, 2022
June 4, 2022
June 1, 2022
April 25, 2022
February 23, 2022
February 6, 2022
December 1, 2021