Data Analysis & Correspondence Analysis
Completed:
Problem Statement
Complex categorical datasets often hide relationships between variables that aren’t immediately visible through simple summary statistics. Understanding these multidimensional patterns requires rigorous statistical testing and visualization techniques to extract actionable insights.
Approach
- Statistical Testing: Performed χ² (chi-square) tests to assess independence between categorical variables
- Dimensional Reduction: Applied Correspondence Analysis to visualize high-dimensional categorical relationships in 2D space
- Data Exploration: Conducted comprehensive exploratory data analysis to identify patterns and anomalies
- Visualization: Created publication-quality charts and heatmaps to communicate findings
Tech Stack
Python Pandas NumPy SciPy Matplotlib Statistics Data Analysis
Key Outcome
Developed a reusable analytical framework for categorical data exploration, successfully uncovering meaningful patterns and relationships that supported data-driven decision-making.
GitHub Repository
Year: 2024
Status: Completed