Skip to content

GB-AFS

GB-AFS (Graph-Based Automatic Feature Selection) is an approach designed to identify the optimal subset of features necessary for maintaining predictive performance, without necessitating user-specified parameters, such as the desired number of features to include. This self-sufficiency is what attributes the 'Automatic' aspect to its name. Operating as a filter-based methodology, GB-AFS is model-agnostic, allowing for the integration of feature selection seamlessly into the preprocessing phase, regardless of the predictive model being used.

The primary innovation and strength of GB-AFS lie in its unique capability to autonomously determine the smallest set of features required, circumventing the common limitation among filter-based methods that typically rely on user input for configuration.

Using GB-AFS: Code Examples and Visualization

GBFS offers a versatile and user-friendly Python library for feature selection in multi-class classification tasks. This section guides you through initializing the GB-AFS object with your dataset, selecting features, and visualizing the feature space.

Initialization and Parameters

To start using GB-AFS, you first need to initialize the GB-AFS object with your dataset and selection criteria:

main.py
1
2
3
4
5
6
7
8
from gbfs import GBAFS

gbafs = GBAFS(
    dataset_path="path/to/your/dataset.csv",
    separability_metric="your_separability_metric",
    dim_reducer_model="your_dimensionality_reduction_method",
    label_column="class",
)

Parameters Explained

  • dataset_path: Path to your dataset file. Ensure your dataset is in a CSV format or another compatible format.
  • separability_metric: Metric for evaluating feature separability.
  • dim_reducer_model: Dimensionality reduction model applying to your dataset. Must implement a fit_transform method for compatibility.
  • label_column: Name of the column with labels in your dataset. Defaults to 'class'.

Current supported metrics for separability_metric are jm, bhattacharyya, and wasserstein. To request support for additional metrics, please open an issue in the repository.

Feature Selection

Once the GB-AFS object is initialized, you can proceed with the feature selection process:

main.py
1
2
3
selected_features = gbafs.select_features()

print("Selected Feature Indices:", selected_features)
This method returns a list of indices for the features deemed most relevant by the GB-AFS algorithm.

Visualizing the Feature Space

GB-AFS also includes a method to visualize the selected features within the feature space, providing insights into their distribution and separability:

main.py
1
gbafs.plot_feature_space()

This method generates a scatter plot highlighting the selected features. Features are displayed with their separability power indicated by color intensity, and selected features are marked distinctly.

References and Further Reading

For a deeper understanding of the GB-AFS method and its background, consider exploring the official paper.