Personal
Taylor Swift Song Data Analysis
- Extracted data from the Spotify API using Python scripting to gather comprehensive song attributes, collecting data for over 500 tracks.
- Cleaned raw data by handling missing values, removing duplicates, and refining the dataset, improving data quality and accuracy by 62%.
- Performed exploratory data analysis and created meaningful visualizations using Matplotlib to identify trends in Taylor Swift’s music, highlighting key patterns across 10+ albums.
Click here to check out the GitHub repository for this project.
Professional
Quantifying Vehicle Steering Discrepancies
Contribution:
- Restructured and normalized over 500 JSON log file records using Pandas to enable structured analysis of steering behavior and identify discrepancies, improving data accessibility for deeper analysis.
- Employed Matplotlib to visualize steering angles and vehicle trajectories, facilitating the visual identification of steering discrepancies.
- Utilized feature engineering to optimize a logistic regression model designed for predicting steering discrepancies, successfully achieving an accuracy rate of 93.4%.
Proactive Road Construction Detection System
Contribution:
- Leveraged BeautifulSoup to web scrape construction-related data from Department of Transportation (DOT) websites across states, ensuring real-time access to critical information.
- Visualized vehicle trajectory and signal data using Matplotlib to identify patterns associated with erratic steering behavior in autonomous vehicles, analyzing data from 50+ vehicle tests to uncover key contributing factors.
- Optimized a logistic regression model to predict erratic steering behavior through feature engineering, achieving an accuracy rate of 93.4%, enhancing predictive reliability.