No1 at the SKU-level in the M5 forecasting competition - Lecture 5.0


00:01 Introduction
01:56 The M5 uncertainty challenge - Data (13)
04:52 The M5 uncertainty challenge - Rules (23)
08:30 The M5 uncertainty challenge - Results (33)
11:59 The story so far
14:56 What is (probably) about to happen
15:43 Pinball loss - Foundation 13
20:45 Negative binomial - Foundation 23
24:04 Innovation Space State Model (ISSM) - Foundation 33
31:36 Sales structure - The REMT model 13
37:02 Putting it together - The REMT model 23
39:10 Aggregated levels - The REMT model 33
43:11 Single stage learning - Discussion 14
45:37 Pattern complete - Discussion 24
49:05 Missing patterns - Discussion 34
53:20 Limits of the M5 - Discussion 44
56:46 Conclusion
59:27 5.0 No1 at the SKU-level in the M5 - Questions?

Description

In 2020, a team at Lokad achieved No5 over 909 competing teams at the M5, a worldwide forecasting competition. However, at the SKU aggregation level, those forecasts landed No1. Demand forecasting is of primary importance for supply chain. The approach adopted in this competition proved to be atypical, and unlike the other methods adopted by the other top 50 contenders. There are multiple lessons to be learned from this achievement as a prelude to tackle further predictive challenges for supply chain.

References

  • A white-boxed ISSM approach to estimate uncertainty distributions of Walmart sales, Rafael de Rezende, Katharina Egert, Ignacio Marin, Guilherme Thompson, December 2021 (link)
  • The M5 Uncertainty competition: Results, findings and conclusions, Spyros Makridakis, Evangelos Spiliotis, Vassilis Assimakopoulos, Zhi Chen, November 2020 (link)