A flexible forecasting model for production systems

TL;DR

This paper introduces a versatile forecasting model tailored for production systems, emphasizing customization, interpretability, robustness, and scalability, with features like automatic changepoint detection and multi-scale seasonality handling.

Contribution

It develops a new family of models that satisfy key properties for production forecasting, including flexibility, interpretability, and speed, with novel automatic changepoint detection capabilities.

Findings

Model captures complex seasonal patterns and events.

Provides robust and scalable forecasting results.

Includes automatic detection of trend and seasonal change points.

Abstract

This paper discusses desirable properties of forecasting models in production systems. It then develops a family of models which are designed to satisfy these properties: highly customizable to capture complex patterns; accommodates a large variety of objectives; has interpretable components; produces robust results; has automatic changepoint detection for trend and seasonality; and runs fast -- making it a good choice for reliable and scalable production systems. The model allows for seasonality at various time scales, events/holidays, and change points in trend and seasonality. The volatility is fitted separately to maintain flexibility and speed and is allowed to be a function of specified features.