metaforecast 0.1.5

Meta-learning and Data-centric Forecasting

metaforecast

Meta-learning and data-centric techniques for time series forecasting models. Built on top of Nixtla's ecosystem, leveraging its state-of-the-art forecasting methods.

Features

metaforecast currently consists of three main modules:

1. Dynamic Ensembles: Leveraging multiple models with adaptive ensemble techniques.
2. Synthetic Time Series Generation: Creating realistic synthetic time series data for robust model training and testing. Includes a special callback for online data augmentation.
3. Long-Horizon Meta-Learning: Instance-based meta-learning for multi-step forecasting.

⚠️ WARNING

metaforecast is in the early stages of development. The codebase may undergo significant changes. If you encounter any issues, please report them in GitHub Issues

Installation

You can install metaforecast using pip:

pip install metaforecast

Documentation

TDA

Examples

Check out the notebooks folder (currently under construction) for example usage and tutorials.