Intro

This page lists some research project ideas that Dr. Ye has been, is or will be working on. Student researchers are always welcome. Email me or come to my office hours. Let’s talk.

You can be a great researcher, if you have some soft skills in you: self-initiative, good communication skill, creativity, dare to dream, can scaffold your way up there, have insight in understanding new material toward that dream, and dare to implement that dream while failing constantly. If you are not sure if your soft skill is up for it and anxious about whether if you can actually do research, some easier-to-measure hard skills: good programming background (proficient with one language, and can design and build a relatively big scale project from ground up) and open to learn new programming languages/tools, and have the time and energy to invest in your own growing, on time for deadlines, can plan work ahead, reply email promptly, know what question to ask, and where to find the answer.

Students with strong technical skills are highly welcome. You are not afraid of challenges and challenges are afraid of you! You don’t have to know every detail in programming, but you know where to ask and what to use. For example, you don’t know remember how to write a merge sort but you can ask ChatGPT to write one for you, and you can tell whether it’s correct or not. You can program a lot faster with the aid of AI.

Students with interdisciplinary background are highly welcome. You may be able to find a special technique that is well-established in AI, and apply it to solve a long-standing issue in your field!

Projects in their embryonic states

Combing GAN and classifier, to generate escher arts.

measuring the open mindedness of AI

Why shouldn’t we use AI for counseling psychology (Collaborating with Dr. Kuo Deng)

Intertwining symbolic AI and statistical AI.

Topology of a neural network.

Collaboration with Animal Science Department. Something involving evolutionary algorithm/ genetic programming. Maybe a herd modeling problem.

Talked with Dr. James Burgess. measuring quality of milk. Similarity measure using ML.

working on

Ongoing projects

NN-kNN, a machine learning model with duality: it can be interpreted as a k-NN, and it can be trained/used like a neural network. Therefore it is powerful and interpretable at the same time.

XAI: expainable AI.

Case-based reasoning (CBR), symbolic reasoning in AI systems

Finished Projects that can be extended

Class-to-class methodology: Study the similarity and difference patterns between classes. (collaboration with Indiana University Bloomington)

Using alternating optimization to improve the 4 stages of CBR.

Technically, anything already done can be extended. link