AI with Community Partners:
This project will use machine learning and AI techniques to assist several community partners with their data driven projects. The mathematics will include learning about and implementing various ML and statistical learning methods like neural nets, random forests, nonnegative matrix and tensor factorizations, and others. We will also develop and design new methods that address the practical needs of our partners that are not currently satisfied by existing approaches. We will carefully consider issues surrounding bias in all of our methods. Our partners include: the California Innocence Project - a nonprofit legal team that works to free innocent people in prison, Homeboy Industries - the largest gang rehabilitation and re-entry program in the world,and CDTech / YLEAD a partner of Public Allies whose mission is to build livable and economically viable communities in the low-income areas of Greater Los Angeles. An example of such projects includes for example a Covid Outreach project with CDTech that aims to further inform and protect those communities through information and resource delivery. In addition to our mathematical contribution to these nonprofits, this team will also participate in outreach activities, meeting with the constitutes and those they serve; in particular to encourage and help prepare them for higher education.
Gang Reduction
We plan to analyze data collected by the city of Los Angeles as part of it's gang reduction program. This data involves both a youth program and a crime reduction program. Recent work in this area by REU students includes natural language processing of text data and dynamic mode decomposition to study the evolution of the program using survey data.
Here are some recent REU papers on this topic:
Marc Andrew Choi, Siyu Huang, Hengyuan Qi, Marco Scialanga, Emerson McMullen, Axel Sanchez Moreno, Yifei Lou, Andrea L. Bertozzi, and P. Jeffrey Brantingham, Combining Dynamic Mode Decomposition and Difference-in-Differences in an Analysis of At-Risk Youth, 2022 IEEE International Conference on Big Data Workshop on Data science for equality, inclusion and well-being challenges, 2022.
Jiaoying Ren, Karina Santoso, David Hyde, Andrea L Bertozzi, P Jeffrey Brantingham, The pandemic did not interrupt LA's violence interrupters, Journal of Aggression, Conflict and Peace Research, dec 12, 2022
Ruofei Wu, Chenxin Yang, David Hyde, Andrea L. Bertozzi and P. Jeffrey Brantingham, Emotion Classification and Textual Clustering Techniques for Gang Intervention Data, workshop on Data Science for Smart and Connected Communities, Proc. Proc. IEEE International Conference on BIG DATA, pp. 3246-3254, 2020.
Sian Wen, Andy Chen, Tanishq Bhatia, Nicholas Liskij, David Hyde, Andrea L. Bertozzi, and P. Jeffrey Brantingham, Analyzing Effectiveness of Gang Interventions using Koopman Operator Theory, workshop on Data Science for Smart and Connected Communities, Proc. IEEE International Conference on BIG DATA, pp. 3237-3245, 2020.
Particle Laden Flow
This is a laboratory based project involving measurement and modeling of slurries. Students will learn laboratory skills and how to take and record data. Students will also work with conservation law models from nonlinear PDE so some background in partial differential equations is desired.
Here are some past REU-involved papers on this topic:
Dominic Diaz, Jessica Bojorquez, Joshua Crasto, Margaret Koulikova, Tameez Latib, Aviva Prins, Andrew Shapiro, Clover Ye, David Arnold, Claudia Falcon, Michael R. Lindstrom, Andrea L. Bertozzi, Investigation of Constant Volume and Constant Flux Initial Conditions on Bidensity Particle-Laden Slurries on an Incline, American J. Undergraduate Research, 16(3), pp. 43-54, 2019.
N. Murisic, J. Ho, V. Hu, P. Latterman, T. Koch, K. Lin, M. Mata, and A. Bertozzi, Particle-laden viscous thin-film flows on an incline: experiments compared with an equilibrium theory based on shear-induced migration and particle settling, Physica D, 240(20), pp. 1661-1673, 2011,
Title: Machine Learning and Nonlinear Dynamics
This project combines neural networks and scientific modeling. Students will work on constructing and applying neural networks toproblems in computational physics. Background: The main requirement is knowledge and/or experience in Python(knowledge in PyTorch or TensorFlow would be helpful). The mathematical requirement is successful completion of a course in ordinary differential equations, e.g. Math upper division differential equations.
Predictors of Gun Niolence using Knowledge Graphs
This is an empirical project on using Twitter/social media data to predict triggers of emotional distress and likelihood of random physical/gun violence. This project will parallel some prior REU projects involving Twitter data:
Clay Adams, Malvina, Bozhidarova, James Chen, Andrew Gao, Zhengtong Liu, J. Hunter Priniski, Junyuan Lin, Rishi Sonthalia, Andrea L. Bertozzi, and P. Jeffrey Brantingham, Knowledge Graphs of the QAnon Twitter Network, Graph Techniques for Adversarial Activity Analytics (GTA3) workshop in IEEE BIG DATA, 2022.
D. Flocco, B. Palmer-Toy, R. Wang, H. Zhu, R. Sonthalia, J. Lin, A. L. Bertozzi, and P. J. Brantingham, An Analysis of COVID-19 Knowledge Graph Construction and Applications, GTA3 workshop, IEEE BIG DATA conference, p. 2631-2640, 2021.
Opinion Dynamics on Networks
This project considers the dynamics of opinions on networks including the spread of ideas and information on social network structures.