Getting started

Getting started#

Here, we’ll show you how to install AutoEmulate and get started with the basics.

✨ Why AutoEmulate?#

  • 🛠️ Diverse Emulators: From classic Radial Basis Functions to cutting-edge Neural Processes

  • 🪄 Low-Code: Data-processing, model comparison, cross-validation, hyperparameter search and more in few lines of code

  • 🎯 Optimized for Emulation: Optimized for typical emulation scenarios with small to medium datasets (100s-1000s of points) with many inputs and outputs

  • 🔌 Easy Integration: All emulators are scikit-learn compatible, and the underlying PyTorch models can be extracted for custom use

  • 🔮 Downstream Applications: Still early days, but we’ve got prediction, sensitivity analysis, history matching and more

🎓 State-of-the-Art Models#

📈 Classical

  • Radial Basis Functions

  • Second Order Polynomials

🌳 Machine Learning

  • Random Forests

  • Gradient Boosting

  • Support Vector Machines

  • LightGBM

Deep Learning

  • Multi-output / Multi-task Gaussian Processes

  • Conditional Neural Processes

Contents