🍹

Cocktail MOBO

Multi-Objective Bayesian Optimization
for the perfect cocktail recipe

1. Choose Ingredients

Add at least 2 ingredients you'll use in your cocktail.

2. Name Your Objectives

Rate cocktails on 3 dimensions (e.g., Sweetness, Sourness, Bitterness).

①
②
③

3. Set Number of Rounds

Exploration (Sobol) samples the space quasi-randomly. Optimization (MOBO) uses a surrogate model to focus on promising recipes.

Total: 25 rounds

Iteration 0 / 25 Sobol Phase
Sobol (exploration) MOBO (optimization)

Try This Mix!

Round 1

Rate This Cocktail

Taste the cocktail and rate each dimension

🏆

Optimization Complete!

All 25 rounds finished. Here are your optimal recipes.

Optimal Recipe (Pareto Average)

This recipe is the average of all Pareto-optimal cocktails found.

Objective Space — Final Pareto Front

Complete view of all 25 tastings. Drag to rotate · Scroll to zoom.

All Pareto-Optimal Recipes

Full Iteration History