TA: Big Data in Transportation

As the Teaching Assistant for Big Data in Transportation (CIEN E4011), I developed and led recitations covering a range of topics essential for handling and analyzing large-scale transportation datasets.

My work included creating and presenting tutorials on:

• High-Performance Computing: Introducing students to JAX for efficient numerical computation, including Just-In-Time (JIT) compilation and automatic parallelization with vmap and pmap.
• Cloud Data and APIs: Demonstrating how to query large public datasets like the NYC Taxi dataset using Google BigQuery and how to leverage the Google Maps API for real-world applications.
• Machine Learning Fundamentals: Building practical examples for both classification (MLP on MNIST) and generative models (fine-tuning a diffusion model on CIFAR-10) using modern frameworks like Flax and Hugging Face Diffusers.
• Model Evaluation and Interpretability: Covering key classification and regression metrics with scikit-learn and introducing model explainability techniques using SHAP to help students understand and validate their models.

You can download the tutorial notebooks here:

• Big Data Tutorial: Python, NumPy, JAX, and BigQuery
• Google Maps API & Machine Learning with JAX and Diffusers
• Model Evaluation Metrics and Explainability with SHAP