5. Discusión
5.5 Capacidad de solubilizar fósforo de las bacterias aisladas
If the number of elements in Xt is finite (and sufficiently small),
sir
algorithms are wasteful because there is usually a positive probability that two available particle paths at Stept are identical, i.e. for anym; n2 Kt,
P˚XB1nWtjt
1Wt DX Bm
1Wtjt
1Wt >0: (2.15)
Thediscrete particle filter(
dpf
) introduced by (Fearnhead, 1998; Fearnhead& Clifford, 2003) tackles this problem by propagating particles in a way that reduces the probability on the left hand side in Equation 2.15 to zero. This is done by extending each available particle trajectory once in every possible direction at Stept. To keep the computational cost from growing
exponentially int, the resulting trajectories are then stochastically pruned
in a way that is optimal in the sense that it minimises the variance of the sum of the self-normalised importance weights.
In this subsection, we show that by using particular choices of the kernelsSt 1,Rt 1,Qt, andt, the
dpf
can be viewed as a special caseof the generic
smc
algorithm. To our knowledge, this is a new result. It immediately implies the validity ofcsmc
algorithms, backward sampling, or ancestor sampling (see Section 3.4 in the next chapter), as described in Whiteley, Andrieu and Doucet (2010), for thedpf
. However, thedpf
cannot be viewed as a special case of
sir
due to the dependence in the proposal kernels and the use of a biased resampling scheme. Here, we recall that in the terminology of Definition 2.3, a resampling scheme is termed ‘biased’ if it does not lead to an evenly weighted (i.e. unweighted) set of particles after resampling. We reiterate that any estimates of in- tegrals of the formt.ft/will still be unbiased as long as the resampledparticles are suitably weighted.
Without loss of generality, assume a finite state spaceXt D NK, for
anyt 2Tfor some (usually not too large)K 2N. At thetth step of the
algorithm, we have Nt WD MK ^Kt particles, where M 2 N can be
chosen to control the computational cost of the algorithm. As described below, at Step t, we selectMt WD Nt=K particle trajectories from the
previous step and extend each of them in allK possible directions.
Resampling Scheme. In this case, we do not make use of the auxiliary
kernelsSs 1. The kernelRs 1D zRs 1is then given by z Rs 1.z1Ws 1;das 1/ D zR?s 1.z1Ws 1;da1sWM1s/ K Y nD2 •a1WMs s 1 .da .n 1/MsC1WnMs s 1 /;
whereRz?s 1.z1Ws 1;da1sWM1s/denotes the resampling scheme (for Ms off-
spring) developed in Fearnhead (1998) which summarised in the following. A more formal description of the entire kernel Rzs 1 can be found in
Section A.5 of the appendix, for completeness.
At Steps, we use Fearnhead (1998, Algorithm 5.2) to solve Ns 1 X nD1 1^Cs 1Wsn1.z1Ws 1/ DMs;
forCs 1 > 0. The idea is that particles whose self-normalised weights
exceed the threshold 1=Cs 1get exactly one offspring. The remaining
particles have at most one offspring.
Collect the indices of the former particles in the set
Ls WD#˚n2Ks 1
ˇ ˇW
n
s 1.z1Ws 1/ >1=Cs 1
and letlsW f1; : : : ;#Lsg !Ls be the function which mapsnto thenth
largest element inLs. We then set the first #Lsparent indices determinist-
ically viaA1W#Ls
s 1 WD.ls.1/; : : : ; ls.#Ls//. The remainingMs #Lsparent
indices take values in Ks 1nLs. They are generated using systematic
resampling based on the weights.Wsn1.z1Ws 1//n2Ks 1nLs, after these have
been re-normalised to sum to 1.
Note thatMs DNs 1impliesCs 11=Œmaxn2Ks 1W
n
s 1.z1Ws 1/and
thusLs DNMs, i.e. in this case, we propagate all existing particle paths
without any pruning.
Proposal Kernel. The proposal kernels are completely deterministic, i.e.
q1.dx1/D•.1;:::;K/.dx1/, and
Qs..z1Ws 1;as 1/;dxs/D•.1Ms;2Ms;:::;KMs/.dxs/;
wheremdenotes anm-component vector of 1s. In other words, each of the Ms particle trajectories chosen as parents by the resampling distribution
2.3 Some Important
smc
AlgorithmsImportance Weights. To ensure that the Radon–Nikodým derivative
x
wt exists, let1.u1; /WD•u1and, fors >1, lets.u1Ws; /be the uniform
distribution onZ.us 1/MsC1;usMs DWD
us
s . Ifsis the time-reversal kernel
from Assumption 2.9, then by the properties of the resampling scheme employed here (see Section A.5 of the appendix),
s..u1Ws;z1Ws 1; bs 1/;fbsg/•bs 1.fa bs s 1g/ D R m s 1..z1Ws 1; bs/;fabss1g/ 1^Cs 1Wsbs11.z1Ws 1/ 1fabs s 1g.bs 1/1Dsus.bs/:
Hence, thenth Step-t particle weight,wnt.z1Wt/, can be written as
wnt.z1Wt/D t.fx bn 1Wtjt 1Wt g/ Qt sD2Œ1^Cs 1W bn s 1jt s 1 .z1Ws 1/ Dwbtn 1jt t 1 .z1Wt 1/ t.x bn 1Wt 1jt 1Wt 1 ;fxtng/ 1^Ct 1W bn t 1jt t 1 .z1Wt 1/ ;
for anyz1Wt in the support of Qt andwnt.z1Wt/D0, otherwise.