Traffic modeling and prediction is a field that has been researched and studied for many years. With the introduction of large data sets taken from sensor stations throughout California, and a myriad of other data sources available contributing to traffic metrics collection, opportunities for analysis into traffic modeling and the factors causing traffic are ever expanding.

As part of the UCSD MAS DSE program, two consecutive cohorts of students have applied data science concepts and analysis to the traffic data. The Cohort 1 team laid the groundwork with data processing, transformation, and principal component analysis. The findings from Cohort 1 are included in this repository as a reference. The focus of the report outlined below is on the efforts of Cohort 2.

1. Introduction

• Abstract
• Advisors
• Team
• Tools
• Resources

2. Data

• Data Description
• Data Acquisition
• Data Preparation

3. Exploratory Analysis

• Traffic State Classification
• Traffic Modeling
• Flow Oscillation ("wiggles")

4. Interactive Visualizations

• Oscillation Visualizations ("wiggles")
• Sensor Health