Voting records of U.S. Congress members on key legislative issues, formatted for binary classification tasks. The dataset originates from the UCI Machine Learning Repository and is commonly used for teaching classification algorithms. The specific temporal coverage and number of records are not detailed in the provided metadata.
Use Cases
- Classify a member's party affiliation based on their votes on specific bills using binary vote features.
- Predict a member's vote on a target issue using other recorded votes as features in a classification model.
- Analyze voting coalitions by clustering members based on patterns across binary vote columns.
- Train binary classifiers like decision trees or logistic regression using the 'yea'/'nay' features for each issue.
Strengths
- Dataset is a canonical benchmark from UCI, ensuring widespread use and validation.
- Structured for binary classification, providing clear features and labels for model training.
Limitations
- Unknown row count limits assessment of statistical power and model generalizability.
- Temporal staleness is a risk as the specific Congress years covered are not provided.
- Potential class imbalance in party affiliation or specific votes is not described.
Provenance
- Source
- UCI Machine Learning Repository
- Collection Method
- null
- Time Range
- null
- Freshness
- null
- Geography
- United States federal legislature