README for "A Large-Scale Empirical Study of Aligned Time Series Forecasting"

Overview

This repository contains results for the paper "A Large-Scale Empirical Study of Aligned Time Series Forecasting". This study explores meta-learning and portfolio approaches for model selection in time series forecasting. The repository contains the results of the experiments with 22 local and 13 global models for time series forecasting.

Authors

• Polina Pilyugina - [email protected]
• Svetlana Medvedeva - [email protected]
• Kirill Mosievich - [email protected]
• Ilya Trofimov - [email protected]
• Alina Kostromina - [email protected]
• Dmitry Simakov - [email protected]
• Evgeny Burnaev - [email protected]

Contents

• input_data/: Directory containing metrics for models, considering horizons of 7, 30, 90, and 365 days.
• results/: Directory for saving the results of the experiments.

For further information, please refer to the full paper or contact the authors.