Examples Gallery

Welcome to the fusionlab-learn Examples Gallery!

This section provides a collection of practical examples and tutorials demonstrating how to use the various features, models, and utilities within the fusionlab-learn library. Each example is designed to showcase a specific aspect of the library or a common workflow.

Browse through the examples below to see fusionlab-learn in action.

Gallery Contents:

• Forecasting Examples
  • Basic Point Forecasting with Flexible TemporalFusionTransformer
  • Quantile Forecasting with TFT Variants
  • Point Forecasting with Stricter TFT (Required Inputs)
  • Advanced Forecasting with XTFT
  • XTFT Forecasting with Anomaly Detection
• Anomaly Detection Examples
  • Prerequisites
  • Example 1: LSTM Autoencoder for Anomaly Detection
  • Example 2: Using SequenceAnomalyScoreLayer (Conceptual)
  • Example 3: Using PredictionErrorAnomalyScore
• Plotting & Visualization Gallery
  • Plotting Utilities
  • Forecast Visualization Utilities
  • Visualizing Forecasts with K-Diagram
• Using Command-Line Tools
  • Using the General TFT CLI (tft_cli.py)
  • Running Specific Application Scripts (Example)

Note

The code examples are designed to be illustrative. For production use, ensure thorough testing, validation, and hyperparameter tuning on your specific datasets.

Tips for Navigating the Gallery:

• Start with Basics: If you’re new, the forecasting examples (e.g., basic TFT, quantile forecasting) are a good starting point.

• Explore by Topic: Use the table of contents to find examples related to specific areas like anomaly detection or data preparation.

• Run the Code: Most examples are self-contained or use datasets available through fusionlab.datasets. We encourage you to download and run the code snippets to get hands-on experience.

We hope these examples help you get the most out of fusionlab-learn!