winedarksea/AutoTS

Automated Time Series Forecasting

What it solves

AutoTS 旨在快速部署大規模高精度的時間序列預測。它自動化選擇最佳預測模型、前處理步驟與集成技術的過程，省去時間序列分析中常見的手動反覆試驗。

How it works

此函式庫提供多種預測模型（簡易、統計、機器學習與深度學習）以及超過 30 種時間序列專屬轉換。它使用遺傳演算法執行 AutoML 特徵搜尋，自動找出資料集的最佳模型與前處理組合。支援多變量輸出、機率預測，且可擴展至數十萬條輸入序列。

Who it’s for

需要在大規模時間序列資料上部署高精度預測，且不想手動調校每個模型的資料科學家與分析師。

Highlights

  • Automated Model Selection: Uses genetic algorithms to find the best models and preprocessing steps.
  • Scalability: Capable of handling tens to hundreds of thousands of input series.
  • Flexible Data Formats: Works directly with Pandas Dataframes in both long and wide formats.
  • Ensembling: Features horizontal and mosaic style ensembles to maximize accuracy while maintaining scalability.
  • Probabilistic Forecasting: Supports upper and lower bound forecasts for uncertainty estimation.