AI+Science Hackathon

The Data Science Institute and the Schmidt AI in Science Postdoctoral Fellows program are excited to invite you to the upcoming AI in Science Hackathon at the University of Chicago, where machine learning meets scientific discovery! This event is designed for UChicago undergraduates and graduate students eager to tackle real-world scientific challenges using AI. The hackathon will take place April 15 – 29 on the University of Chicago campus.

Hackathon Highlights

• Two Projects: Competitors will have the opportunity to work on one of two projects. The first project will focus on characterizing new materials using AI, and the second project will focus on using reinforcement learning to control networks of living neurons.
• Team Competition: Each team will have 4 members. Participants can sign up as individuals or as groups of up to 4. Individuals or groups of less than 4 will be matched with others.
• Mentorship: Each team will be assigned a mentor from the Schmidt AI+Science fellows program to provide guidance and support throughout the hackathon.
• Resources: We provide access to NVIDIA A100 GPUs on RCC’s midway cluster to train your models.
• Prize: The winning team of each project will receive a prize.
• Location: Teams will meet in person at the kickoff and final presentation. To get the most out of the Hackathon we highly recommend meeting in person during the Hackathon. We provide rooms on select dates on the UChicago campus to meet and a slack space to coordinate virtually.

First Project Overview

Characterize New Research Materials with AI

We present an exciting opportunity for graph-neural networks (GNNs) to tackle a real-world research problem that could enable future x-ray experiments to instantly identify the elements present in novel research materials. X-rays are not just the workhorse for imaging bone structures in our bodies, they are also the workhorse of materials characterization techniques across the biological, chemical and physical sciences. X-rays interact with the electrons most tightly bound to an atom’s nucleus, which enables them to identify the elements present in new research materials. The central property identified by x-rays is the core-electron binding energy (CEBE). In the 1980s researchers published a database of 2,800 experimental CEBE values which was recently digitized into an online, open-access format. This presents us with an opportunity to use graph-neural networks to predict the CEBE and avoid the costly and complex computation of these values with quantum mechanics, accelerating the development of novel research materials in clean-energy and quantum hardware technologies.

For this project, we will provide teams with the dataset processed into molecular graphs. The teams will then develop a GNN model to predict the CEBE from the molecular structure. This high-impact research challenge, provides an opportunity to use GNN’s on scientific data and unlock the transformative potential offered by AI to under-explored scientific disciplines.

Second Project Overview

Reinforcement Learning for Biological Neuronal Networks

This project explores how we can use artificial intelligence to teach biological neuronal networks in real time. Biological neuronal networks, grown from brain cells in our laboratory, are tiny networks of living neurons that mimic the way our brains process and store information. Traditionally, fixed stimulation patterns have been applied to these networks to study their properties, such as memory capabilities and learning. However, such an approach can’t keep up with the natural changes that occur within the biological network over time, such as neuron growth or shifts in activity. In addition, it cannot adapt to the network under consideration.

By using reinforcement learning (RL) – a type of AI where an agent learns by trial and error – we can create smarter stimulation patterns that influence the network’s behavior in real time. Think of it as a dynamic conversation between the AI and the neurons, where both sides learn and improve as they interact. RL helps us not only better understand how neurons learn and adapt but also opens the door to using these networks for advanced applications, like testing new drugs or building bio-hybrid systems.

For this project, we provide you with biological networks from which the neuronal activity can be recorded. Your goal is to write an RL agent, which electrically stimulates the neuronal cultures. The project will teach you the basics of reinforcement learning, with a focus on sample-efficiency.

Eligibility

• Open to all University of Chicago undergraduate and graduate students. If you are a staff member or postdoctoral fellow and are interested in participating, please contact us directly.

How to Apply

• Please fill out the application form. Every team member needs to sign up individually. We have a limited number of spots available, so sign up as soon as possible!

Application Deadline

• March 28th