This paper proposes using two metrics to quantify the forecastability of time series prior to model development: the spectral predictability score and the largest Lyapunov exponent. Unlike traditional model evaluation metrics, these measures assess the inherent forecastability characteristics of the data before any forecast attempts. 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 alternative strategies for products with limited forecastability.

The outcome of all time series cannot be forecast, e.g. the flipping of a fair coin. Others, like the repeated {01} sequence {010101...} can be forecast exactly. Algorithmic information theory can provide a measure of forecastability that lies between these extremes. The degree of forecastability is a function of only the data. For prediction (or classification) of labeled data, we propose three categories for forecastability: oracle forecastability for predictions that are always exact, precise forecastability for errors up to a bound, and probabilistic forecastability for any other predictions. Examples are given in each case.

Forecasting building energy consumption has become a promising solution in Building Energy Management Systems for energy saving and optimization. Furthermore, it can play an important role in the efficient management of the operation of a smart grid. Different data-driven approaches to forecast the future energy demand of buildings at different scale, and over various time horizons, can be found in the scientific literature, including extensive Machine Learning and Deep Learning approaches. However, the identification of the most accurate forecaster model which can be utilized to predict the energy demand of such a building is still challenging.In this paper, the design and implementation of a data-driven approach to predict how forecastable the future energy demand of a building is, without first utilizing a data-driven forecasting model, is presented. The investigation utilizes a historical electricity consumption time series data set with a half-hour interval that has been collected from a group of residential buildings located in the City of London, United Kingdom

Picture this: You are working on a data science project that requires you to predict the demand of a product. Your manager or stakeholders tell you that if your model's forecasts are within 15% of the true demand, then we'll call this project a success. You can't wait to begin the project. You assemble the dataset, conduct data analysis, and notice that the demand of the product is very erratic with no obvious patterns. Nonetheless, you proceed by cleaning the data, filling in the missing values, and creating features that you think might be valuable for your machine learning model.

You are right that this is a question of forecastability. There have been a few articles on forecastability in the IIF's practitioner-oriented journal Foresight. The problem is that forecastability is already hard to assess in "simple" cases. Suppose you have a time series like this but don't speak German: How would you model the large peak in April, and how would you include this information in any forecasts? Unless you knew that this time series is the sales of eggs in a Swiss supermarket chain, which peaks right before western calendar Easter, you would not have a chance.

I introduce Forecastable Component Analysis (ForeCA), a novel dimension reduction technique for temporally dependent signals. Based on a new forecastability measure, ForeCA finds an optimal transformation to separate a multivariate time series into a forecastable and an orthogonal white noise space. I present a converging algorithm with a fast eigenvector solution. Applications to financial and macroeconomic time series show that ForeCA can successfully discover informative structure, which can be used for forecasting as well as classification. The R package ForeCA accompanies this work and is publicly available on CRAN.