The Center for Mathematical Artificial Intelligence (CMAI) is a newly established research center under the Department of Mathematics and Institute of Mathematical Sciences (IMS) in CUHK. For the development and flourishment of forefront aritificial intelligence technologies, we are committed to providing a platform on which various research scientists from different fields and areas can carry out researches with rigorous mathematics.
We are fully devoted to our mission statement:
"We aim to provide a platform for research scientists in academics and industries to carry out rigorous analysis and develop cutting-edge AI technologies through advanced mathematics."
Artificial intelligence (AI) aims to build intelligence through machines for high-level understanding. AI has become more popular in recent years, thanks to increased amount of available data, advanced algorithms, improvements in computing power and storage. It makes it possible for machines to learn from experience, adjust to new inputs, perform and outperform human-like tasks. It can be foreseeable that AI will play a crucial role to build a better world. Nevertheless, the theoretical aspects and scientific computing aspects of many AI techniques are still lacking. Understanding these aspects is necessary for enhancing the effectiveness, efficiency and accuracy of various AI technologies.
Our ultimate goal of establishing the captioned center (CMAI) is to bring together experts and train young research scientists to analyze the theoretical and computational aspects of AI, through rigorous mathematical concepts and tools. New AI methods and corresponding numerical computation techniques are to be developed, which can be practically applied in real industries. In short, CMAI will be a platform for bringing together academics, scientists, and entrepreneurs in AI, who are passionate about innovation and creating a better future, to analyze and develop cutting-edge AI technologies through advanced mathematics.
Various AI algorithms and tools, especially deep learning based methods, have been developed in recent years. However, the theoretical understanding of the methods is lacking. Theoretical analysis is important to better understand the strength and weakness of the methods, so that more effective and efficient new tools can be developed. Our center aims to analyze the theoretical and computational aspects of AI, through rigorous mathematical concepts and tools.
With the theoretical understanding of various AI techniques, we will develop new computational methods, which are more efficient and accurate. We will systematically study the numerical convergence and stability of the developed methods. The developed computational methods are expected to be applied to practical real world problems, such as medical related problems, computer visions problems and so on.
To train young generation to enter field, our center will offer non-degree training courses about some basic techniques related to machine learning. The target audiences are undergraduate students in their 3rd or 4th years with basic mathematical knowledge. Our training courses will cover mathematical concepts necessary for machine learning methods as well as state-of-the-art machine learning algorithms.
To help young research scientists to get involved in AI research, we will organize regular seminar and study group to learn state-of-the-art machine learning technologies. The target participants are PhD students and postdoctoral scholars, who are interested in AI. The research seminar aims to provide a platform for young research scientists to learn recent published papers related to AI. The study group aims to bring together young researchers to rigorously study new techniques and mathematical methods related to AI.
AI research is not limited to the field of mathematics but multi-disciplinary. Our center will explore collaboration amongst various departments/schools in CUHK. One potential collaboration is with the CUHK medical school to develop medical image processing and analysis tools based on deep neural network methods.
To bring about research impact, we aim to establish collaboration with international industrial partners. We will invite researchers from industries to visit our center from time to time to share the real-world problems they are solving. Our goal is to contribute with our mathematical models to build AI tools to solve practical real-world problems to benefit the society. It will promote knowledge transfer and research impact.
To better share research ideas, we will regularly organize international conferences to invite international research scientists in the fields to present their latest researches and exchange research ideas. These conferences can also help to promote research collaboration amongst international researchers from all over the world.