Zhizhuo Yang Email and Phone Number

CSCI-635 Introduction to Machine Learning. Topics include: Mathematical background of machine learning (e.g. statistical analysis and visualization of data), neural models (e.g. Convolutional Neural Networks, Recurrent Neural Networks), probabilistic graphical models (e.g. Bayesian networks, Markov models), and reinforcement learning.

• Developed a machine learning pipeline for data storage, processing, training, and evaluation using cloud infrastructures.• Pre-trained models on user data and fine-tuned for customized tracking abilities, reducing mean tracking accuracy by 51% and variance by 32%.• Researched gradient-based EKF learning for sensor fusion and developed a neural network-based residual pose learning algorithm.• Processed 1.35 TB of data using Python, C++, and Bash scripts; implemented methods in PyTorch, reducing mean tracking error by 39.76%.• Implemented methods in PyTorch and conducted training on a high-performance computing cluster, achieving a 39.76% reduction in mean tracking error and a 12.68% decrease in standard deviation.

• Developed a real-time statistical filtering algorithm in C++ for improving Asynchronous TimeWarp technology in VR, reducing visual artifacts.

CSCI-620 Intro to big data course covers data storage, data cleaning, data mining etc. It touches databases such as MySQL and MongoDB and clustering algorithms.

• Created a deep learning-based algorithm for frame extrapolation to enhance rendering frame rates and save GPU resources on VR headsets.• Implemented and trained models using TensorFlow and PyTorch with Horovod and Slurm.

Teaching assistant for graduate course: Intro to Big Data (CSCI-620).