P ′ x (x, p) = A =
A 11 0 0 A 22
, P ′ p (x, p) = H(x) =
x 1 0 1 0 0 x 2 0 1
(3.4.9)
where
n p = 2n
. The new redued matrixH
is the result of eliminating the seond and the third olumn of the oldH
, sine we have eliminated theseondandthirdelement of
p
. Itisstraightforward toprodue thematriesA
andH
orresponding to an arbitrary element elimination of the full pa-rametervetorp ∈ R (n(n+1))
.When using the diagonal matrix approah we redue the number of input
parameters by a fator
n+1
2
. Aordinglywe an expet a redution of thealulationtime, whihan beuseful inproblems oflarge dimension.
A dierent approah ould be to alter the number of parameters for every
mainiteration,withtheaimof thepossibilityofrepeatedlyhavingaunique
solution to the Parameter Extration. The length of the residual vetor is
inreased by
1
for eah main iteration, starting atn
fork = 0
, and we anaugment the
p
-vetor with one extra element. The order of the elementsan be hosen in many ways, a suggestion would be to begin with
p = [α]
and augment with the diagonal elements of
A
one by one followed by theb
-elements. The initial value for the mapping parameters would still beA (0) i = I
,b (0) i = 0
,α (0) i = 1
,and this initial valueis valid, until theparam-eteris ontained in
p
.lem, if an exat math is possible in the ases, where we have either more
parameters than residual elements or the same number of both. Usually
it is possible to satisfy all equations exatly, but one an nd examples of
problems,where itisnot. Weassumethattheproblemsonsideredhereare
onsistent. In ase of a full rank onsistent problem the weighting of the
residual elements has no eet on thesolution. When theproblem is rank
deient the solution spae for the unweighted problem is the same as for
the weighted problem, andthe twoproblems have idential minimum norm
solutions.
Inthe aseof an non-onsistent problem, theweighting fatorsan havean
eet.
Underdetermined System of Equations
Iftherearemoreparameters
n p
than residualelementsn r
,theproblemdoesnot have a unique solution. The Jaobian of
r
is denotedJ r
and has therank
q
, whereq ≤ n r
. There will then be an(n p − q)
-innity of solutionsto the Parameter Extration. For eah of the innitely many solutions we
arethen sure,that all
n r
equations aresatised,ie.r(p ∗ ) = 0
. Algorithm3provides the solutiondened bythe Marquardtparameters
µ
.Themultipliationwiththeweigthingfatorsnowresultsintheresidual
Wr
andthe Jaobian
WJ r
,wherer
andJ r
refer toequation (3.3.9) ,andwherethe diagonal matrix
W
isgiven by(3.5.1) :W =
w 1 0
.
.
.
w k σ
.
.
.
0 σ
(3.5.1)
Inaseof a onsistent problem thesolutionto thepenalizedproblem isthe
same as the unpenalized solution. In this ase the penalty fators have no
eet onthesolution.
Overdetermined System of Equations
Provided that
J r
hasfullrankintheasewheren r = n p
,thesolutiontotheParameter Extration isunique. All theequations aresatisedexatly,and
sine
r(p ∗ ) = 0
,thenalsoW · r(p ∗ ) = 0
.Iftherearemoreequationsthanunknownparameters(
n r > n p
),wendtheleastsquaressolutionbyAlgorithm3. Weannotbesuretosatisfyall
equa-tions, and the weight of eah equation determines, whih solution is found,
sine this solution minimizes
1/2 · r
Tr
. A largepenaltyfator will resultina solution thatis guaranteed to satisfythegradient residual, provided that
thesaling of theresidual elements isnot bad. If theproblem onthe other
hand isonsistent,the penaltyfator hasno inuene.
3.6 The Weighting Fators
Sine the weighting fators funtion in the same way as the penalty fator
disusedintheprevious setion,theanalysisregardingthesystemsof
equa-tions isalso valid inthepresent setion.
In the new Spae Mapping formulation we wish the mapping parameters
to minimize the residual funtion (1.3.12) , onsisting of the funtion value
residualandthegradientresidual. Weassume,thatwehavemade
k + 1
iter-ations inthe mainalgorithm. Therstiterateis
x (0)
. Theresidualdependson
k
ofthek + 1
iterates,aswellasthegradient wrt.thedesignparametersinthebestiterate, sinethefuntionvaluesoftheneandsurrogatemodels
in the best iterate already math. The length of the residual vetor is
de-noted
n r
anddependson the numberof main iterationsand thenumber ofdesignparameters (
n r = k + n
). Withuseoftheregularizationtermn r
alsodependsonthenumberofmapping parameters,inthisase
n r = k + n + n p
.We introdue weighting fatorsforthefuntion valueresidualrows denoted
w i
fori = 1, . . . , k
. Thegradientresidualisweightedwiththepenaltyfatorσ
,whihis disussedinsetion 3.5.Linear Weight Funtion
IntheimplementationoftheSMISframeworkbyFrankPedersen,the
weight-ingfatorsaregiven bya linearfuntion depited ingure3.6.1.
where
ε res
is a given threshold value (the gure showsε res = 0.25
). Theweighting funtion dereases linearly from
1
to0
as the distane to thebest iteration point grows, giving the expression for the weighting fators
w 1 , . . . , w k
:0 0.2 0.4 0.6 0.8 1 0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
dX
w
Linear weight function w(dX)
Threshold value ε r = 0.25
Figure3.6.1: Linear weight funtion
w i =
1
fordX i < ε res 2 − dX ε res i
for
dX i < 2ε res
0
fordX i ≥ 2ε res
Gauss Distributed Weight Funtion
Another approah is to let the weighting fators depend on thenumber of
residualrows
n r
. Thenumber ofunknown parameters isn p
,henewe needatleast
n p − n + 1
iterationsto ensure,thattheproblemis not underdeter-mined.We hoosea weight funtion asa Gaussurve.
0 0.2 0.4 0.6 0.8 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
dX
w
Gauss weight function w(dX)
Threshold value ε r = 0.1
Figure3.6.2: Gaussdistributed weight funtion
w i =
( 1
ifn r < n p
exp( − γ · dX i 2 )
ifn r ≥ n p
(3.6.1)
where the fator
γ
is determined byγ = − ln(ε dX ¯ res 2 )
, withdX ¯
orresponding to the(n p − n)
th best iterate. This gives the weighting fatorε res
for thispartiular residualelement. Inthisways theweighting fatorsfor the
resid-ual elements orresponding to the best iterates are kept above or equal to
the threshold value
ε res
, and all other points are onsidered low priority.Theweighting funtion for
ε r = 0.10
isseen ingure3.6.2. Ifthere arenotenough rows in the residual to math the number of unknown parameters,
allresidual rows areweighted equallywithafator
1
.IftheParameterExtrationproblemisoverdetermined,theweightingfators
makesure thatthe bestpointsaregiven the highestpriority. Asmentioned
before the weighting fators do not inuene the solution, if the system is
onsistent. The threshold valueis usedto deide,how important arole the
'unneessary'iterates should playinndingthe optimal parameter set.
If theParameter Extration problem is underdetermined, it is usually
pos-sibleto satisfyall equations exatly. Inthis asethe weighting fatorshave
no eet on the solution, sine if
r(p ∗ ) = 0
then alsoW · r(p ∗ ) = 0
. Thestrategy for determining the weighting fators inthease
n r < n p
isthere-forenot important,aslong asthe