NIMAA

NIMAA performs nominal data mining in R by integrating graph theory and data mining to analyze datasets composed of nominal variables.

Key Features:

• Graph Construction: Creates weighted and unweighted bipartite graphs from nominal variables.
• Label Analysis and Clustering: Computes label similarities and clusters labels into super-labels.
• Validation and Prediction: Validates clustering results and predicts bipartite edges by imputing missing edge weights.
• Visualization Tools: Provides visualization functions for interpreting graph-based structures and clustering outcomes.

Scientific Applications:

• Nominal data exploration in biology: Enables identification of patterns and relationships among categorical variables in complex biological datasets.

Methodology:

Integration of graph-theoretic and data-mining methods to construct weighted and unweighted bipartite graphs from nominal variables; computation of label similarity and clustering into super-labels; validation of clustering and imputation of missing edge weights to predict bipartite edges; generation of visualizations for results.