Mapping Dark Matter with Bayesian Neural Networks w/ Yashar Hezaveh - TWiML Talk #250
Published:Apr 11, 2019 19:01
•1 min read
•Practical AI
Analysis
This article summarizes a discussion with Yashar Hezaveh, an Assistant Professor at the University of Montreal, focusing on his work using machine learning to analyze gravitational lensing. The core of the discussion revolves around applying ML to correct distorted images caused by gravity, specifically in the context of mapping dark matter. The conversation touches upon the integration of simulations and ML for image generation, the use of techniques like domain transfer and GANs, and the methods used to evaluate the project's outcomes. The article highlights the intersection of astrophysics and machine learning, showcasing how AI is being used to solve complex scientific problems.
Key Takeaways
- •The article discusses the application of machine learning, specifically Bayesian Neural Networks, to map dark matter.
- •The research utilizes gravitational lensing, the bending of light due to gravity, to analyze distant sources.
- •The project involves using ML to correct distorted images and incorporates techniques like domain transfer and GANs.
Reference
“Yashar and I discuss how ML can be applied to undistort images, the intertwined roles of simulation and ML in generating images, incorporating other techniques such as domain transfer or GANs, and how he assesses the results of this project.”