Coherent forecast combination for linearly constrained multiple time series

arXiv:2412.03429 · stat.ME, stat.AP, stat.CO · Submitted 2024-12-04 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Statistics. The day in statistics: methods, theory and the inference under them.

Tom: We're looking at the paper: "Coherent forecast combination for linearly constrained multiple time series".

Jane: Here's the gist of the paper. Linearly constrained multiple time series, encountered in contexts like National Accounts or multilevel frameworks, often require methods to combine multiple incoherent base forecasts into a coherent final result to enhance accuracy and achieve coherence.

Tom: This paper develops an optimization-based technique for coherent forecast combination that combines multiple unbiased base forecasts while respecting the constraints of the time series.

Jane: The authors derive closed-form expressions for both the coherent combined forecast vector (y c = psi y b = M omega y b, where M = I n - W c C C C-1, omega = W-1 K c, and W c = K W-1 K-1) and its error covariance matrix (W fc = M W c M).

Tom: The theoretical foundation relies on the assumptions that base forecasts are unbiased and the error covariance matrix is known, with practical estimation methods like using in-sample residuals outlined.

Paper summary: Tom: So we're looking at this paper now titled "Coherent forecast combination for linearly constrained multiple time series". This whole thing is about taking several separate, sometimes messy forecasts and putting them together in a way that respects the rules of how those variables are linked.

Jane: Exactly. The core idea here is that when you have multiple forecasts for things that are all related by some kind of constraint—like how different parts of an economy fit together—you can get a much better final prediction by combining them coherently instead of just picking the best one or averaging them up blindly.

Lu: What the authors are doing is developing an optimization-based technique to do this combination while making sure the resulting forecast still respects those original constraints, which is what makes it coherent. They aim to combine multiple unbiased base forecasts without violating those time series rules.

Meng: From my side, I wonder how robust this method is when we're talking about real-world data. They claim they've derived closed-form expressions for both the combined forecast vector and its error covariance matrix, which sounds mathematically neat on paper but I need to know what that means for deployment.

Lalam: The paper presents closed form expressions for the coherent combined forecast vector, which is written as y sub c equals Psi transpose y sub b equals M omega transpose y sub b, where M is a matrix defined as one minus W c C C transpose C inverse, and omega is defined as W inverse K c, and W c is equal to K transpose W inverse K inverse >

Tom: That sounds like some heavy linear algebra. So what's the big picture here for us? Why should someone listening on their drive care about this kind of mathematical trick with time series?

Jane: It matters because many real-world problems, like national accounts or energy consumption data, are structured hierarchically. If you only look at one part, it's incomplete. This paper suggests that using all the available forecasts together can improve the accuracy and make the final result more sensible for decision-making.

Lu: The theoretical foundation of this work rests on two main assumptions: first, that all your base forecasts are unbiased, meaning they aren't systematically over- or under-predicting anything, and second, that you already know the error covariance matrix for those base forecasts.

Meng: Knowing the error covariance matrix is a big assumption in practice. How do you get those matrices reliably from real traffic data? What if the structure of that error isn't exactly what the model assumes?

Lalam: The paper suggests using in-sample residuals as a practical way to estimate that error covariance matrix, which is how they make their theoretical framework usable in a setting where we don't know everything beforehand >

Paper summary: Tom: Okay, so they have this theory and some practical estimation methods. What did they actually test? Did they just build this thing and say it works?

Jane: They did empirical validation on the daily Australian electricity generation hierarchical time series. They compared their coherent combination approach against several other things, like individual base forecasts, single-task combinations, and even sequential coherent combinations.

Lu: The results show that the coherent multi-task forecast combination consistently outperformed all those benchmark methods across different settings for both upper and bottom time series >

Meng: Outperformed how many? I need to know the actual performance gain. Are we talking about a small improvement in mean square error, or something substantial enough to justify the complexity of this optimization technique?

Tom: The paper specifically points out that the optimal combination forecast vector, which minimizes the mean square error, is given by y equals Psi transpose y b equals M omega transpose y b >

Jane: That specific combination formula is what they found to be best for minimizing error when you're looking for an unbiased estimator within all possible linear combinations of your base forecasts >

Lu: The study looked at various performance metrics, including the Average Relative Mean Square Error and the Average Relative Mean Absolute Error >

Meng: I see those tables. They compare different forecast horizons and panel sizes. What are some of the caveats they mention about these comparisons? Is this method only good for one specific structure of constraints?

Lalam: They show results across different forecast horizons, panel sizes, and aggregation levels for both upper and bottom time series >

Tom: It seems like the coherent approach, which they call occbe in their simulations on the Australian data, beats methods like ew which is a common benchmark >

Jane: The conclusion of this work is that the problem boils down to finding the best unbiased estimator you can get by linearly combining all your base forecasts, and their method, occbe, came out as the most effective overall technique they tested >

Lu: They are also flagging that future work will focus on more advanced covariance matrix estimation and extending this framework into probabilistic forecasting scenarios >

Meng: So for someone trying to deploy this in production right now, what's the immediate practical hurdle? Is it the computation cost of solving that optimization problem, or is it just trusting the closed-form expressions they gave us?

Tom: The paper doesn't really go into heavy computation cost analysis, but if you have a huge number of variables, solving that optimization might become slow. That's something we'll need to watch >

Jane: It sounds like this is a solid theoretical step forward for combining forecasts in complex systems, moving beyond just picking one model >

Lu: It moves the conversation from simply comparing models to finding the mathematically optimal way to combine multiple uncertain inputs while respecting structural constraints >

Conclusion: Tom: So this paper is about taking several different forecasts for related things—like parts of an economy—and figuring out the best way to stitch them together so the final result makes sense given all those rules.

Jane: It’s titled "Coherent forecast combination for linearly constrained multiple time series," and it’s from some researchers looking at how to combine data when you have these tight structural links between variables.

Lu: The authors are using an optimization method to find that best stitch, aiming to combine forecasts without breaking those original time series constraints.

Meng: From my side, the big question is whether this combination actually holds up when we try to use it for real-time predictions where we can't just look at perfect historical data.

Lalam: I’m looking at the literature and I’m pretty sure the core idea here is that finding this optimal combination is about finding the best unbiased estimator you can get from all your separate base forecasts.

Tom: Finding that best unbiased estimator sounds like it could be really useful if we're dealing with complex systems where a single forecast just isn't enough.

Jane: Exactly, because instead of trusting just one model, this method suggests we can use the information from every available base forecast to get a more accurate picture.

Lu: The paper shows that their method, which they call occbe in their tests on Australian electricity data, actually performs better than just using individual forecasts or combining them sequentially.

Meng: Better how much? Is this just a tiny bit of improvement over what we’re currently doing with simple averaging?

Tom: The authors show it consistently outperforms benchmarks like ew across different time horizons and panel sizes in their simulations on that electricity data set.

Jane: It suggests that for complex, constrained systems, the way you combine the inputs matters just as much as the individual inputs themselves.

Lu: So they are pointing toward finding this optimal linear combination formula to handle those structural dependencies correctly.

Meng: It sounds like a solid theoretical step forward, but I still need to see how computationally expensive solving that optimization actually is when we scale it up to our massive production models.

University of Padua

stat.ME, stat.AP, stat.CO

Submitted: 2024-12-04

Updated: 2026-11-06

Code: https://github.com/PuwasalaG/Probabilistic-Forecast-Reconciliation

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 77/100

The gist: Linearly constrained multiple time series, encountered in contexts like National Accounts or multilevel frameworks…

Key concepts

Coherent Forecast Combination
This is the process of merging several separate, potentially conflicting base forecasts into one single, accurate final forecast. The goal is to create a result that respects the underlying constraints of the time series data while minimizing prediction errors.
Linearly Constrained Multiple Time Series
These are multiple time series (like different economic indicators) that must follow specific linear relationships or constraints. This structure is common in complex systems like national accounts, where one variable's forecast depends on others.
Optimal Unbiased Estimator
The study frames the problem as finding the best possible way to combine all available base forecasts into a single estimate that remains unbiased. The authors conclude that this optimal combination method, called occbe, is the most effective overall approach found.

Terminology

Summary

Linearly constrained multiple time series, encountered in contexts like National Accounts or multilevel frameworks, often require methods to combine multiple incoherent base forecasts into a coherent final result to enhance accuracy and achieve coherence. p. 0

Where the paper says it

Coherent forecast combination for linearly constrained multiple time series Daniele Girolimetto∗ , Tommaso Di Fonzo Department of Statistical Sciences, University of Padua, Padova 35121, Italy Abstract Linearly constrained multiple time series may be encountered in many practical contexts, such as the National Accounts (e. , GDP disaggregated by Income, Expenditure and Output), and multilevel frameworks where the variables are organized according to hierarchies or groupings, like the total energy consumption of a country disaggregated by region and energy sources.

This paper develops an optimization-based technique for coherent forecast combination that combines multiple unbiased base forecasts while respecting the constraints of the time series. p. 0
Where the paper says it

In these cases, when multiple incoherent base forecasts for each individual variable are available, a forecast combinationand-reconciliation approach, that we call coherent forecast combination, may be used to improve the accuracy of the base forecasts and achieve coherence in the final result. In this paper, we develop an optimization-based technique that combines multiple unbiased base forecasts while assuring the constraints valid for the series.

The authors derive closed-form expressions for both the coherent combined forecast vector (y c = psi y b = M omega y b, where M = I n - W c C C C-1, omega = W-1 K c, and W c = K W-1 K-1) and its error covariance matrix (W fc = M W c M). p. 0
Where the paper says it

In this paper, we develop an optimization-based technique that combines multiple unbiased base forecasts while assuring the constraints valid for the series. We present closed form expressions for the coherent combined forecast vector and its error covariance matrix in the general case where a different number of forecasts is available for each variable.

The theoretical foundation relies on the assumptions that base forecasts are unbiased and the error covariance matrix is known, with practical estimation methods like using in-sample residuals outlined. p. 0

Where the paper says it

In this paper, we develop an optimization-based technique that combines multiple unbiased base forecasts while assuring the constraints valid for the series. We present closed form expressions for the coherent combined forecast vector and its error covariance matrix in the general case where a different number of forecasts is available for each variable.

The empirical validation, conducted on the daily Australian electricity generation hierarchical time series, demonstrates that this coherent multi-task forecast combination consistently outperforms individual base forecasts, single-task combinations, reconciliation by single experts, and sequential coherent combination approaches. p. 0
Where the paper says it

Through simulations and a forecasting experiment on the daily Australian electricity generation hierarchical time series, we show that the proposed methodology, in addition to adhering to sound statistical principles, may yield in significant improvement on base forecasts, single-task combination and single-expert reconciliation approaches as well. Keywords: Forecasting; Linearly constrained multiple time series; Coherent forecasts; Forecast combination; Forecast reconciliation; Australian electricity generation ∗Corresponding author Email address: daniele.

Specifically, the optimal (minimum mean square error, MMSE) coherent linear combination forecast vector is given by y e = psi y b = M omega y b. p. 4
Where the paper says it

scr Sequential first-combination-then-reconciliation. occ Optimal (minimum mean square error, MMSE) coherent linear combination.

The study comprehensively evaluates various reconciliation and combination methods. p. 1

Where the paper says it

It is important to note that the generation of the initial base forecasts is outside the scope of this study, as these forecasts may be produced through various methodologies. The classical univariate forecast combination approach (Bates and Granger, 1969, Clemen, 1989, Timmermann, 2006, Wang et al.

The analysis utilizes multiple performance metrics, including Average Relative MSE (Table E. p. 2
Where the paper says it

Section 3 describes the model linking the multiple base forecasts to the target linearly constrained forecast vector. The methodological analysis, including the closed-form solution for optimal coherent forecast combination and its covariance matrix, is presented in Section 4.

16, E.18), Average Relative MAE (Table E. p. 9
Where the paper says it

A possible, practical solution to obtain coherent forecasts is adopting a two-step procedure (sequential combination-first-then-reconciliation, scr), consisting of performing the reconciliation of yb c in order to finalize the coherent combination forecasting process (see Rostami-Tabar and Hyndman, 2024). Another simple two-step procedure, sequential reconciliation-first-then-simple-average-combination, src, may be developed as well.

17), and statistical tests such as the MCB Nemenyi test (Appendix F) and pairwise DM-tests based on absolute loss and quadratic loss (Figure G. p. 4
Where the paper says it

yn,t⊤ ∈ R n be a vector of observed time series we are interested in forecasting, and assume that the time series is linearly constrained, in the sense that at every time t the n individual variables of yt must satisfy nu < n independent linear constraints. Without loss of generality, we assume that the nu constraints are expressed in zero-constrained form, according to the homogeneous linear system: Cyt = 0(nu×1), (1) 5

6). The comparison spans different forecast horizons, panel sizes, and aggregation levels for both upper and bottom time series, detailed in Tables G. p. 6
Where the paper says it

(2011), the structural form (2) is widely used in the literature on forecast reconciliation for genuine hierarchical/grouped time series (for a recent review, see Athanasopoulos et al. In fact, expression (2) has the merit of describing the relationships linking the free (bottom) and the constrained (upper) time series in a way that visually recalls a genuine hierarchical or grouped structure, by placing the aggregated series at higher levels than the bottom one, where the most disaggregated series take place.

1 through G.3, which present AvgRelMAE (Table G. p. 0
Where the paper says it

In this paper, we develop an optimization-based technique that combines multiple unbiased base forecasts while assuring the constraints valid for the series. We present closed form expressions for the coherent combined forecast vector and its error covariance matrix in the general case where a different number of forecasts is available for each variable.

19) and AvgRelMSE (Table G. p. 3
Where the paper says it

For matrices A, A ⪰ 0 means that A is positive semi-definite. Table 1 summarizes the notation used throughout the paper.

20), along with Model Confidence Set results using absolute loss (Table G. p. 2
Where the paper says it

In Section 2 we set the notation and consider different representatons of the problem. Section 3 describes the model linking the multiple base forecasts to the target linearly constrained forecast vector.

21) and quadratic loss (Table G. p. 4
Where the paper says it

yn,t⊤ ∈ R n be a vector of observed time series we are interested in forecasting, and assume that the time series is linearly constrained, in the sense that at every time t the n individual variables of yt must satisfy nu < n independent linear constraints. Without loss of generality, we assume that the nu constraints are expressed in zero-constrained form, according to the homogeneous linear system: Cyt = 0(nu×1), (1) 5

22). The results consistently show that coherent approaches, such as occbe, outperform benchmark methods like ew in simulations on the Australian daily electricity generation time series. p. 0
Where the paper says it

Through simulations and a forecasting experiment on the daily Australian electricity generation hierarchical time series, we show that the proposed methodology, in addition to adhering to sound statistical principles, may yield in significant improvement on base forecasts, single-task combination and single-expert reconciliation approaches as well. Keywords: Forecasting; Linearly constrained multiple time series; Coherent forecasts; Forecast combination; Forecast reconciliation; Australian electricity generation ∗Corresponding author Email address: daniele.

The authors conclude that the problem is best specified as finding the optimal unbiased estimator within the class of all unbiased estimators obtainable via linearly combining all available base forecasts, with occbe emerging as the most effective overall method. Future work will focus on more advanced covariance matrix estimation and extending the framework to probabilistic forecasting. p. 2
Where the paper says it

Section 5 discusses practical issues related to the estimation of the covariance matrix. Insights into the empirical performance of the newly proposed methodology are presented in Section 6, that describes a Monte-Carlo experiment, and in Section 7, where an application on forecasting the daily time series of the Australian electricity generation is discussed.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Optimization-based technique for coherent forecast combination: The system can solve a linearly constrained quadratic program to find an optimal coherent combined forecast vector, as described in Theorem 1, which minimizes the mean square error while ensuring constraints are satisfied.

  2. Covariance matrix estimation using in-sample residuals: The system can estimate the error covariance matrix W by using in-sample forecast errors (residuals) from training data, such as the sample forecast MSE matrix or shrunk versions thereof, providing a practical path for practitioners when the covariance matrix is not known.

  3. Structural approach equivalence: The system can derive optimal coherent linear forecast combination results using either the zero-constrained approach or the structural approach, as these two methods are shown to yield equivalent results for unbiased forecasts.

  4. Forecast accuracy evaluation via model confidence sets: The system can identify statistically superior forecasting approaches by utilizing the Model Confidence Set (MCS) approach developed by Hansen et al., which helps in selecting approaches with statistically superior performance across different forecast horizons and time series.

  5. Performance comparison of coherent combination methods: The system can outperform individual base forecasts, single-task combinations, and single-expert reconciliation approaches, specifically noting that the optimal coherent combination approach occbe is the most effective overall in terms of both mean square error and mean absolute error.

Abstract

Forecasting linearly constrained multiple time series, as those arising in National Accounts, energy systems and supply chain planning, requires that the forecasts of all variables jointly satisfy the underlying accounting or resource-balance identities. When several experts provide forecasts for some or all of the variables, classical forecast combination operates on each variable and ignores the linear constraints, while forecast reconciliation enforces such constraints by using a single forecast per variable. To bridge this gap, we develop an original coherent forecast combination framework, in which the joint problem of combination across experts and coherence across variables is formulated as a linearly constrained optimization problem. Closed-form expressions are derived for the coherent combined forecast and its error covariance matrix, and the statistical properties of the resulting estimators are discussed. A Monte Carlo experiment and a forecasting application to an Australian daily electricity generation dataset confirm that the proposed approach significantly improves forecast accuracy with respect to standard forecast combination and reconciliation alternatives.

Related papers