Frequently Asked Questions and forecasting tips

Thank you for choosing skforecast and visiting our Frequently Asked Questions (FAQ) page. Here, we aim to provide solutions to commonly encountered issues on our Github repository. If your question is not answered on this page, we encourage you to create a new issue on our Github so that it can be addressed, and other users can also benefit from it. Additionally, we have included some forecasting tips to help you get the most out of skforecast.

General Forecasting Tips

• Avoid negative predictions when forecasting
• Forecasting time series with missing values
• Forecasting time series with trend: differentiation
• Forecasting time series with trend: tree based models
• Forecasting with delayed historical data
• Backtesting vs One-step-ahead
• Continuous Ranked Probability Score (CRPS)
• Calibration of probabilistic forecasting intervals
• Avoid data leakage in pre-trained forecasting models

Feature Engineering

• Cyclical features in time series
• Time series aggregation

Performance Optimization

• Parallelization in skforecast
• Profiling skforecast