报告题目：Machine Learning Constraints on Dark Matter from 21cm Signals

报告人：Kenji Kadota 副研究员（国科大杭州高等研究院）

报告摘要：We explore the application of convolutional neural networks (CNNs) to 21cm maps as a tool for probing dark matter properties. Following a review of 21cm physics and CNN algorithms, we discuss three concrete dark matter scenarios: ultra-light dark matter (axion-like particles), annihilating dark matter, and warm dark matter.

We demonstrate the utility of CNNs in differentiating dark matter models by linking 21cm map features to the underlying physics of dark matter. Our study highlights image-based cosmology as a novel framework for studying dark matter.

报告时间：2024年12月12日（周四）14:00

报告地点：仙林园区3-302室

欢迎大家参加！