Supervisor: Wei-Chih Huang
Project description
As we have known, artificial intelligence (AI) has been ubiquitous with numerous applications and has resulted in far-reaching influences in our daily lives. Due to the availability of a tremendous amount of digital data, deep learning is a booming branch in AI, which has developed a large variety of neural networks and algorithms.
Deep neural networks (DNNs), which can reproduce arbitrary functions given enough numbers of neurons and layers, are efficient at discovering underlying patterns and correlations among the input and output parameters. Out of many types of DNNs, Convolutional Neural Network (CNN) is renowned for the capabilities of discovering patterns, shapes and features for input images.
The project aims to explore various applications of neural networks to dark matter physics. For instance, one can utilize the neural networks to classify different dark matter density profiles based on the rotation curves. Alternatively, neural networks can be employed to quickly infer experimental constraints on the dark matter interactions.