Multiple PhD positions in NeuroSymbolic programming at Rutgers University

We are looking for motivated Ph.D. students to join the Rutgers Automated Reasoning group!

Our research lies at the intersection of automated reasoning and machine learning, namely neurosymbolic programming. A neurosymbolic program is like a normal program in a compositional, high-level programming language, except it is also allowed to invoke a set of neural networks as library routines. Because it contains high-level, compositional elements, such a learning representation is more human-interpretable, and more easily analyzed using symbolic tools from programming languages and formal methods. While learning these programs, we can go beyond data; we can also provide a strong inductive bias for the learner to encode rich domain knowledge using programming constructs (e.g. loops to capture repetitive behaviors). Doing so can lead to more reliable and generalizable learning. Specifically, we are investigating how to apply neural program synthesis and formal program reasoning to make neurosymbolic systems more reliable and trustworthy. The long-term research goal is to build intelligent and interpretable learning systems that allow the tight integration of deep learning and symbolic reasoning and that can be certified robust and reliable.
Please drop me an email (hz375 AT cs.rutgers.edu) with your CV if you are interested. The application deadline is Jan. 1st. 2022.

罗格斯大学靠近纽约，交通便利，美食丰富，相信会是同学们求学的好地方：）

请问老师对学生的背景、技能有什么要求或期望呢？

織川秀雄 发表于 2021-12-21 19:59
请问老师对学生的背景、技能有什么要求或期望呢？

谢谢你的问题。如果有reinforcement learning, adversarial training, 或者program synthesis的研究背景，会是很大的加成。