Time Series Forecastability Measures

Rui Wang, Steven Klee, Alexis Roos

Keywords: Time Series, Supply Chain, Forecasting

Abstract:

This paper proposes using two metrics to quantify the forecasta-
bility of time series prior to model development: the spectral pre-
dictability score and the largest Lyapunov exponent. Unlike tradi-
tional model evaluation metrics, these measures assess the inherent
forecastability characteristics of the data before any forecast at-
tempts. The spectral predictability score evaluates the strength and
regularity of frequency components in the time series, whereas the
Lyapunov exponents quantify the chaos and stability of the system
generating the data. We evaluated the effectiveness of these metrics
on both synthetic and real-world time series from the M5 forecast
competition dataset. Our results demonstrate that these two metrics
can correctly reflect the inherent forecastability of a time series
and have a strong correlation with the actual forecast performance
of various models. By understanding the inherent forecastability
of time series before model training, practitioners can focus their
planning efforts on products and supply chain levels that are more
forecastable, while setting appropriate expectations or seeking al-
ternative strategies for products with limited forecastability.

Submission Number: 6