Welcome to BoTier Documentation

Next to the “primary” optimization objectives, optimization problems often contain a series of subordinate objectives, which can be expressed as preferences over either the outputs of an experiment, or the experiment inputs (e.g. to minimize the experimental cost). BoTier provides a flexible framework to express hierarchical user preferences over both experiment inputs and outputs.

BoTier is a lightweight plug-in for BoTorch, and can be readily integrated with the BoTorch ecosystem for Bayesian Optimization.

Getting started

BoTier in Brief

What is BoTier and how can it help you?

How to Install

A step-by-step guide to installing BoTier and setting it up for your data processing tasks.

Usage

BoTier Tutorial

Learn how to create and run your first BoTier optimization campaign.

API Reference

Explore the documentation for each submodule and its functionalities.