Tree-based method

As explained in Demand Forecasting Best Practices

A tree-based machine learning model makes predictions using one or more decision-tree structures that represent relationships between inputs and outputs.

A tree-based method is a machine-learning approach that learns relationships between input data and desired outputs, then uses those relationships to forecast demand. It is especially useful when planners need to identify which demand drivers matter, while recognizing that the model does not reveal relationships as clearly as traditional forecasting components. section 15.2 section 15.5

Use tree-based methods with these practical expectations:

• Inspect feature importance to identify potentially irrelevant demand drivers, but do not treat importance as the size or direction of an effect. section 15.1.2
• Run contrasting forecast scenarios, such as promotions versus no promotions, to estimate business effects that feature importance alone cannot explain. section 15.1.2
• Treat the model as a limited-visibility system: it will not directly expose level, trend, or seasonality in the way component-based forecasts can. section 15.1.2

Promotion scenarios

A demand planner compares forecasts with and without discounts. The scenarios indicate a 17% total uplift, while also showing lower sales before and after promotion periods because of self-cannibalization. section 15.1.2

Figure 15.2 Feature importance

Use the feature-importance figure to see how tree-based models rank demand features and why ranking does not establish causal impact. section 15.1.2

How Demand Forecasting Best Practices covers it

How deep it goes

Explains feature importance and scenario analysis, but stops short of exposing level, trend, seasonality, or direct causal effects.

Examples

Uses feature rankings, weather as a low-importance driver, and promotion scenarios showing uplift and self-cannibalization.

What it assumes

Assumes attention to objectives, metrics, and data, while treating supervised learning inputs and desired outputs as the working foundation.

What you can do after

Evaluate demand drivers and compare forecast scenarios while avoiding causal claims from feature importance alone.

How it connects

Connects supervised learning, demand forecasting, feature importance, scenario analysis, and neural networks.

liveBookmanning.com

Read Demand Forecasting Best Practices on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Machine Learning for Tabular Data: XGBoost, Deep Learning, and AIA tree-based model is a predictive model whose decision structure is formed from one or more decision trees.
• 

  liveBookmanning.com
  Machine Learning Algorithms in DepthA tree-based model makes predictions by recursively partitioning an input space into regions represented by nodes and leaves.

Compare how 3 books explain tree-based method →

Related concepts

• random forestincludes · A random forest is an ensemble model that combines predictions from multiple decision trees built with randomized data samples, feature selections, or both.
• importance scoreused by · An importance score is a numerical value indicating an item's relative significance, relevance, or contribution within a specified context.
• XGBoostincludes · XGBoost is an open-source software library for efficient gradient-boosted decision-tree algorithms used in supervised machine learning.
• gradient boostingincludes · Gradient boosting is an ensemble-learning technique that sequentially adds predictive models, typically weak learners, to reduce a loss by fitting its gradient.
• decision treeincludes · A decision tree is a branching structure that represents sequential tests or choices, with paths leading to outcomes, predictions, or actions.
• label encodingused by · Label encoding is a categorical-data transformation that maps each category to a numerical code, which may imply an ordering among categories.

This explainer was generated by Manning's knowledge graph from the text of Demand Forecasting Best Practices. The section links point to the parts of the book it draws on.