Catherine Weaver

Using advanced model-based controls, machine learning, and reinforcement learning to control the world around us.

About Me

Professional Experience

Research Projects

- Modular Control

- End-to-end Learning

- Publications

Service & Teaching

Resume



Contact

catherine22@berkeley.edu

📄

Summary

  • Future Ph.D. from UC Berkeley developing state-of-the-art machine learning and control algorithms for continuous time systems
  • Experience pushing the limits of artificial intelligence with high-quality, highly-collaborative code at Sony AI
  • Over two years of cumulative experience in industry in control systems, computer science, product development, and manufacturing

📰

News

  • Feb 2024: My co-first author paper "Skill-Critic: Refining Learned Skills for Reinforcement Learning" was accepted for publication in IEEE's Robotic and Automation Letters.
  • Jan 2024: My first author paper "Real-time Trajectory Generation via Imitation Learning of Dynamic Movement Primitives for Autonomous Racing" was accepted to the 2024 American Control Conference.

🎓

Education

Ph.D. Mechanical Engineering
University of California, Berkeley | May 2024
  • PI: Masayoshi Tomizuka
  • Minors: AI & Optimization
M.S. Mechanical Engineering
University of California, Berkeley | May 2021
B.S. Mechanical Engineering
Purdue University | May 2019
  • GPA 3.98 - Highest Distinction
  • Spanish Language Minor

⌨️

Programming

Python 🔵🔵🔵🔵🔵
PyTorch 🔵🔵🔵🔵🔵
Tensorflow 🔵🔵🔵🔵⚪
C/C++ 🔵🔵🔵🔵⚪
MATLAB 🔵🔵🔵🔵🔵
Simulink 🔵🔵🔵🔵⚪
Excel VBA 🔵🔵🔵🔵⚪
LaTex 🔵🔵🔵🔵🔵
HTML 🔵🔵🔵⚪⚪

💻

Select Algorithms

  • Model Predictive Control (MPC, iLQR)
  • Reinforcement Learning (SAC, PPO)
  • Imitation Learning (GANs, GAIL)
  • Time-series segmentation (HDPHMM)
  • Sequence Modeling (Decision Transformer)
  • Attention Mechanisms (VAE)
  • Hierarchical RL (Options, Skills, Goals

Skills

  • Environment control (conda, Docker)
  • Version control (Github)
  • Coding style guidelines and typing (MyPy)
  • Cloud computing (AWS)
  • GPU training (CUDA)