Efficient State-space Exploration in Massively Parallel Simulation Based Inference

06/29/2021
by   Sourabh Kulkarni, et al.
0

Simulation-based Inference (SBI) is a widely used set of algorithms to learn the parameters of complex scientific simulation models. While primarily run on CPUs in HPC clusters, these algorithms have been shown to scale in performance when developed to be run on massively parallel architectures such as GPUs. While parallelizing existing SBI algorithms provides us with performance gains, this might not be the most efficient way to utilize the achieved parallelism. This work proposes a new algorithm, that builds on an existing SBI method - Approximate Bayesian Computation with Sequential Monte Carlo(ABC-SMC). This new algorithm is designed to utilize the parallelism not only for performance gain, but also toward qualitative benefits in the learnt parameters. The key idea is to replace the notion of a single 'step-size' hyperparameter, which governs how the state space of parameters is explored during learning, with step-sizes sampled from a tuned Beta distribution. This allows this new ABC-SMC algorithm to more efficiently explore the state-space of the parameters being learnt. We test the effectiveness of the proposed algorithm to learn parameters for an epidemiology model running on a Tesla T4 GPU. Compared to the parallelized state-of-the-art SBI algorithm, we get similar quality results in ∼ 100x fewer simulations and observe  80x lower run-to-run variance across 10 independent trials.

READ FULL TEXT

page 3

page 7

research
09/07/2018

Scalable Monte Carlo inference for state-space models

We present an original simulation-based method to estimate likelihood ra...
research
03/08/2018

Accelerating a fluvial incision and landscape evolution model with parallelism

Solving inverse problems and achieving statistical rigour in landscape e...
research
05/05/2023

Designing Proposal Distributions for Particle Filters using Integrated Nested Laplace Approximation

State-space models are used to describe and analyse dynamical systems. T...
research
08/04/2020

GenCos' Behaviors Modeling Based on Q Learning Improved by Dichotomy

Q learning is widely used to simulate the behaviors of generation compan...
research
11/30/2017

State Space LSTM Models with Particle MCMC Inference

Long Short-Term Memory (LSTM) is one of the most powerful sequence model...
research
03/28/2015

A Multi-signal Variant for the GPU-based Parallelization of Growing Self-Organizing Networks

Among the many possible approaches for the parallelization of self-organ...
research
11/23/2011

Self-Avoiding Random Dynamics on Integer Complex Systems

This paper introduces a new specialized algorithm for equilibrium Monte ...

Please sign up or login with your details

Forgot password? Click here to reset