Skip to main content
xplainabledocs
Version: v1.3.0

Classification – Multi-Class

Coming Soon

Multi-Class Classification is currently in active development and will be available in an upcoming release of xplainable. We're working hard to bring you transparent, explainable multi-class models with the same ease of use you've come to expect.

What to Expect​

When released, xplainable's Multi-Class Classification will provide:

🎯 Multi-Class Support

Handle classification problems with 3+ classes while maintaining full transparency and explainability.

🔍 Class-Specific Insights

Understand what drives predictions for each individual class with detailed feature importance.

⚡ Real-Time Explanations

Get instant explanations for multi-class predictions with the same speed as binary classification.

🎨 GUI Integration

Train and explore multi-class models using the intuitive xplainable GUI interface.

Planned Features​

XMultiClassifier API​

The upcoming XMultiClassifier will follow the same intuitive API pattern as our binary classifier:

from xplainable.core.models import XMultiClassifier

# Simple, familiar API
model = XMultiClassifier()
model.fit(X_train, y_train)
predictions = model.predict(X_test)

# Get explanations for each class
explanations = model.explain(X_test)

GUI Integration​

Train multi-class models with the embedded GUI:

import xplainable as xp

# Initialize session
xp.initialise(api_key=os.environ['XP_API_KEY'])

# Train with GUI (coming soon)
model = xp.multiclass_classifier(data)

Partitioned Multi-Class Models​

Support for partitioned multi-class models for complex segmentation:

from xplainable.core.models import PartitionedMultiClassifier

# Advanced partitioning (coming soon)
partitioned_model = PartitionedMultiClassifier(partition_on='segment')

Current Alternatives​

While we work on multi-class support, you can:

1. Use Binary Classification​

For problems with 3+ classes, consider:

  • One-vs-Rest approach: Train separate binary classifiers for each class
  • Binary decomposition: Break down into multiple binary problems

2. Preprocessing Strategies​

  • Class grouping: Combine similar classes into broader categories
  • Hierarchical classification: Use a tree-like structure of binary classifiers

3. Stay Updated​

  • Follow our releases: Check the GitHub repository for updates
  • Join our community: Get notified when multi-class support is released

Timeline​

Development Status

Multi-class classification is a high priority feature currently in active development. We're targeting release in the coming months and will announce availability through our official channels.

Get Notified​

Want to be the first to know when multi-class classification is available?

  • ⭐ Star our GitHub repository
  • 📧 Follow our release notes
  • 💬 Join our community discussions

In the meantime, explore our powerful binary classification and regression capabilities, or check out advanced topics for sophisticated modeling techniques.