Bayesian Computation Methods for Inference in Stochastic Kinetic Models

Algorithm 1

Particle MCMC algorithm targeting [6].

Initialization  ():

1. Sample and

2. run a SMC scheme targeting . Draw from the SMC approximation and let

denote the marginal likelihood estimate.

Iteration  ():

1. Sample and

2. run a SMC scheme targeting . Draw , let denote the marginal likelihood estimate,

and

3. with probability

accept the move setting , and . Otherwise store the current values ,

and .