Gradient Boosting Application in Forecasting of Performance Indicators Values for Measuring the Efficiency of Promotions in FMCG Retail

TL;DR

A new approach to forecasting promotion efficiency is proposed, using the gradient boosting method for this task, with six performance indicators introduced to capture the promotion effect.

Abstract

In the paper, a problem of forecasting promotion efficiency is raised. The authors propose a new approach, using the gradient boosting method for this task. Six performance indicators are introduced to capture the promotion effect. For each of them, within predefined groups of products, a model was trained. A description of using these models for forecasting and optimising promotion efficiency is provided. Data preparation and hyperparameters tuning processes are also described. The experiments were performed for three groups of products from a large grocery company.

Figures (3)

• Figure 1: Finding matching record without promotion
• Figure 2: Flowchart of the hyperparameter optimisation process.
• Figure 3: Plot of feature importance for the model of the indicator AVG. AMOUNT for dairy products. most important features are shown and the important values are represented as relative to the highest ranked feature.