POETs

POETs (Pareto Optimal Ensemble Technique) is a software tool implemented in the Julia programming language (JuPOETs) for estimating parameter or model ensembles using multiobjective optimization. It combines simulated annealing with Pareto optimality to find ensemble solutions that balance multiple, often conflicting, training objectives. JuPOETs is particularly useful for addressing model uncertainty in deterministic mathematical models by considering parameter or model families instead of single best-fit parameters or fixed model structures.

Key features of JuPOETs include:

1. Ability to handle various problems, such as mixed binary and continuous variable types, bilevel optimization problems, and constrained problems without modifying the core algorithm.

2. Faster performance compared to a similar implementation in Octave, with approximately six-fold speedup on a suite of test functions.

3. Successful application to a proof-of-concept biochemical model with four conflicting training objectives, producing an ensemble of parameters that simultaneously fit the mean of the training data and performed well on individual objective functions.