Effective Bayesian Modeling of Groups of Related Count Time Series

By Nicolas Chapados. 2014

33 citations recorded by the source corpus.

Time series of counts arise in a variety of forecasting applications, for which traditional models are generally inappropriate. This paper introduces a hierarchical Bayesian formulation applicable to count time series that can easily account for explanatory variables and share statistical strength across groups of related time series. We derive an efficient approximate inference technique, and illustrate its performance on a number of datasets from supply chain planning.

Subfields: Artificial Intelligence, Probabilistic & Bayesian Methods, Time-Series & Forecasting

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Probabilistic & Bayesian Methods

In Time-Series & Forecasting

Citations in the corpus

DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks (2017, methodology): “onditions, forecasting methods have also incorporated more suitable likelihood functions, such as the zero-inﬂated Poisson distribution, the negative binomial distribution [20], a combination of both [4], or a tailored multi-stage likelihood [19]. Sharing information across time series can improve t”
Enhancing the Locality and Breaking the Memory Bottleneck of Transformer on Time Series Forecasting (2019, methodology): “[18] proposes hierarchical Bayesian methods to learn across multiple related count time series from the perspective of graph model.”
Approximate Bayesian Inference in Linear State Space Models for Intermittent Demand Forecasting at Scale (2017, methodology): “illa implementation would scale cubically). While previously used in Statistics [8, Sect. 10.7], this reduction is not widely known in Machine Learning. If L-BFGS [14] is used instead (as proposed in [7]), approximate inference fails in our real-world use cases. A multi-stage likelihood, tailored to”
Deep Learning for Time Series Forecasting: Tutorial and Literature Survey (2020, background): “In an early example of hierarchical Bayesian models, [31] combined global and local features for intermittent demand forecasting in retail planning.”
GluonTS: Probabilistic Time Series Models in Python (2019, background): “Classical models that were developed for these tasks (Hyndman and Athanasopoulos, 2017), are, with some exceptions (Chapados, 2014), local models that learn the parameters Φ for each time series individually.”
Deep State Space Models for Time Series Forecasting (2018, background): “This includes monthly and quarterly time series from the tourism competition dataset [2] describing tourism demand, hourly time series from the M4 competition [20] and parts dataset [6] which contains monthly demand of spare parts at a US auto-mobile company.”