Question 1

How to handle continuous variables in a Node.js decision tree?

Accepted Answer

Enable autoDetectTypes in configuration to automatically use the CART algorithm for continuous features, with options for thresholds and statistical tests. The README provides code examples for regression and mixed data.

Question 2

Decision Tree vs Random Forest vs XGBoost in JavaScript – which one should I use?

Accepted Answer

Use Decision Tree for speed and interpretability, Random Forest for balanced performance and reduced overfitting, and XGBoost for highest accuracy on complex data, as compared in the algorithm table with training times and use cases.

Question 3

How to export and import trained models for saving?

Accepted Answer

Call toJSON() to export models as JSON, and use new DecisionTree(modelJson) or the import() method to reload them. This allows persistence without retraining, though exported models don't store original training data.

Question 4

What Node.js version is required for this machine learning library?

Accepted Answer

It requires Node.js 20+ or Bun 1.0+ due to ES module support and modern JavaScript features, as specified in the installation section. CommonJS is not supported.

Question 5

How does this compare to Python's scikit-learn for decision trees?

Accepted Answer

While scikit-learn offers more algorithms and better performance for large datasets, this library provides native JavaScript/TypeScript integration, automatic type detection, and easier deployment in Node.js without external dependencies.

Question 6

Can I use it for regression tasks with continuous targets?

Accepted Answer

Yes, it supports regression via the CART algorithm and XGBoost with objectives like 'regression', using criteria such as MSE for continuous value prediction, as shown in the regression examples.

Decision Trees

What is Decision Trees?

Overview

Use Cases

Best For

Related Projects

Found a gem we're missing?

Not Ideal For

Pros & Cons

Pros

Cons

Frequently Asked Questions