Milestone 1

Traffic Pattern Analysis: Milestone 1 Report

Project Objectives: In Milestone 1, we achieved significant progress toward our goals:

Data Cleaning: Successfully completed the cleaning and organization of raw data, ensuring it's ready for in-depth analysis. This process involved removing inconsistencies and errors from the data to ensure accuracy.
Data Visualization: Implemented visualizations, particularly plotting X and Y coordinates, to represent individual tracks for unique recording IDs. Visualization is crucial for understanding complex data patterns.
Clustering Algorithm: Implemented Dynamic Time Warping (DTW), a technique that calculates an optimal match between two given sequences (e.g., time series). We used DTW for precise track comparison, a vital step in our analysis.

Open-source dataset (InD dataset)

Challenges and Lessons Learned: Through our work, we realized the critical importance of understanding the direction of traffic flow. This insight highlighted the need to consider various variables such as X Acceleration (rate of change of acceleration in the X direction), X Velocity (speed in the X direction), Y Acceleration (rate of change of acceleration in the Y direction), Y Velocity (speed in the Y direction), Frame (specific time frames), and Heading (direction of the vehicle). With a plethora of data at our disposal, selecting the right features for our machine-learning algorithms proved to be a significant challenge. We navigated this challenge to ensure the accuracy and relevance of our analyses.

Next Phase Objectives: Moving forward, our focus will be on:

Traffic Flow Analysis: Implementing machine learning models for in-depth traffic flow analysis. This involves studying the movement patterns of vehicles to understand traffic dynamics better.
Feature Selection: Carefully selecting features (variables) for training our machine learning algorithms. Proper feature selection is crucial for the accuracy of predictive models.
Data Analysis: Conducting comprehensive analysis to gain deep insights. This includes identifying patterns, trends, and anomalies within the data.
Plotting and Visualization: Developing detailed visualizations to enhance our understanding of the data. Visualization aids in communicating complex data insights effectively.
Outlier Detection: Identifying outliers within the data for further investigation. Outliers are data points that significantly differ from the rest of the data and can provide valuable insights or indicate errors.

Project Tools and Resources:

Data Source: We utilized the extensive InD Dataset, comprising data from 4 intersections, each with 8 sets of data (tracksMeta, tracks, recordingMeta).
Visualization: Visual representations of the data were created to aid in our analysis and understanding.

IT Capstone Milestone 1 Report_all

Page updated

Google Sites

Report abuse