# A wall-time minimizing parallelization strategy for approximate Bayesian computation

## TL;DR

This paper introduces a new parallelization strategy for ABC methods that reduces computing time by avoiding idle resources.

## Contribution

The novel contribution is a theoretical proof of unbiasedness for preemptive sampling in ABC parallelization.

## Key findings

- Look-ahead scheduling reduces wall-time by 10-20% and up to 50% compared to existing methods.
- The strategy is compatible with adaptive distance and summary statistic selection.
- It improves resource utilization on high-performance computing infrastructure.

## Abstract

Approximate Bayesian Computation (ABC) is a widely applicable and popular approach to estimating unknown parameters of mechanistic models. As ABC analyses are computationally expensive, parallelization on high-performance infrastructure is often necessary. However, the existing parallelization strategies leave computing resources unused at times and thus do not optimally leverage them yet. We present look-ahead scheduling, a wall-time minimizing parallelization strategy for ABC Sequential Monte Carlo algorithms, which avoids idle times of computing units by preemptive sampling of subsequent generations. This allows to utilize all available resources. The strategy can be integrated with e.g. adaptive distance function and summary statistic selection schemes, which is essential in practice. Our key contribution is the theoretical assessment of the strategy of preemptive sampling and the proof of unbiasedness. Complementary, we provide an implementation and evaluate the strategy on different problems and numbers of parallel cores, showing speed-ups of typically 10-20% and up to 50% compared to the best established approach, with some variability. Thus, the proposed strategy allows to improve the cost and run-time efficiency of ABC methods on high-performance infrastructure.