BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Estimates for distributions of Hölder semi-norms of random proces
ses from spaces F_ψ(Ω)
DTSTART;VALUE=DATE-TIME:20170816T123000Z
DTEND;VALUE=DATE-TIME:20170816T130000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-270@indico.uu.se
DESCRIPTION:Speakers: Dmytro Zatula (Taras Shevchenko National University
of Kyiv)\nIn the following we deal with estimates for distributions of Hö
lder semi-norms of sample functions of random processes from spaces $\\mat
hbb{F}_\\psi(\\Omega)$\, defined on a compact metric space and on an infin
ite interval $[0\,\\infty)$\, i.e. probabilities\n$$\\mathsf{P}\\left\\{\\
sup\\limits_{\\substack{0x\\right\\}.$$\nSuch estimates and assumptions un
der which semi-norms of sample functions of random processes from spaces $
\\mathbb{F}_\\psi(\\Omega)$\, defined on a compact space\, satisfy the Hö
lder condition were obtained by Kozachenko and Zatula (2015). Similar resu
lts were provided for Gaussian processes\, defined on a compact space\, by
Dudley (1973). Kozachenko (1985) generalized Dudley's results for random
processes belonging to Orlicz spaces\, see also Buldygin and Kozachenko (2
000). Marcus and Rosen (2008) obtained $L^p$ moduli of continuity for a wi
de class of continuous Gaussian processes. Kozachenko et al. (2011) studie
d the Lipschitz continuity of generalized sub-Gaussian processes and provi
ded estimates for the distribution of Lipschitz norms of such processes. B
ut all these problems were not considered yet for processes\, defined on a
n infinite interval.\n\nhttps://indico.uu.se/event/317/contributions/270/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/270/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Modeling of vertical and horizontal variation in multivariate func
tional data
DTSTART;VALUE=DATE-TIME:20170815T140000Z
DTEND;VALUE=DATE-TIME:20170815T143000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-271@indico.uu.se
DESCRIPTION:Speakers: Niels Olsen (Københvans Universitet)\nWe present a
model for multivariate functional data that simultaneously model vertical
and horisontal variation. \nHorisontal variation is modeled using warping
functions represented by latent gaussian variables.\nVertical variation is
modeled using Gaussian processes using a generally applicable low-paramet
ric covariance structure.\nWe devise a method for maximum likelihood esti
mation using a Laplace approximation and apply it to three different data
sets.\n\nhttps://indico.uu.se/event/317/contributions/271/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/271/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Finite Mixture of C-vines for Complex Dependence
DTSTART;VALUE=DATE-TIME:20170817T070000Z
DTEND;VALUE=DATE-TIME:20170817T073000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-272@indico.uu.se
DESCRIPTION:Speakers: O. Ozan Evkaya (Atılım University)\nRecently\, the
re has been an increasing interest on the combination of copulas with a fi
nite mixture model. Such a framework is useful to reveal the hidden depend
ence patterns observed for random variables flexibly in terms of statistic
al modeling. The combination of vine copulas incorporated into a finite mi
xture model is also beneficial for capturing hidden structures on a multiv
ariate data set. In this respect\, the main goal of this study is extendin
g the study of Kim et al. (2013) with different scenarios. For this reason
\, finite mixture of C-vine is proposed for multivariate data with differe
nt dependence structures. The performance of the proposed model has been t
ested by different simulated data set including various tail dependence pr
operties.\n\nhttps://indico.uu.se/event/317/contributions/272/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/272/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Testing independence for multivariate time series by the auto-dist
ance correlation matrix
DTSTART;VALUE=DATE-TIME:20170815T093000Z
DTEND;VALUE=DATE-TIME:20170815T100000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-273@indico.uu.se
DESCRIPTION:Speakers: Maria Pitsillou (Department of Mathematics &\; St
atistics\, Cyprus)\nWe introduce the notions of multivariate auto-distance
covariance and correlation functions\nfor time series analysis. These con
cepts have been recently discussed in the context\nof both independent and
dependent data but we extend them in a different direction by\nputting fo
rward their matrix version. Their matrix version allows us to identify pos
sible\ninterrelationships among the components of a multivariate time seri
es. Interpretation\nand consistent estimators of these new concepts are di
scussed. Additionally\, we develop\na test for testing the i.i.d. hypothes
is for multivariate time series data. The resulting test\nstatistic perfor
ms better than the standard multivariate Ljung-Box test statistic. All the
\nabove methodology is included in the R package dCovTS which is briefly i
ntroduced in\nthis talk.\n\nhttps://indico.uu.se/event/317/contributions/2
73/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/273/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Theoretical and simulation results on heavy-tailed fractional Pear
son diffusions
DTSTART;VALUE=DATE-TIME:20170818T083000Z
DTEND;VALUE=DATE-TIME:20170818T090000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-274@indico.uu.se
DESCRIPTION:Speakers: Ivan Papić (Department of Mathematics\, J.J. Stross
mayer University of Osijek)\nWe define heavy-tailed fractional reciprocal
gamma and Fisher-Snedecor diffusions by a non-Markovian time change in the
corresponding Pearson diffusions. We illustrate known theoretical results
regarding these fractional diffusions via simulations.\n\nhttps://indico.
uu.se/event/317/contributions/274/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/274/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Copula based BINAR models with applications
DTSTART;VALUE=DATE-TIME:20170818T090000Z
DTEND;VALUE=DATE-TIME:20170818T093000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-275@indico.uu.se
DESCRIPTION:Speakers: Andrius Buteikis (Faculty of Mathematics and Informa
tics\, Vilnius University)\nIn this paper we study the problem of modellin
g the integer-valued vector observations. We consider the BINAR(1) models
defined via copula-joint innovations. We review different parameter estima
tion methods and analyse estimation methods of the copula dependence param
eter. We also examine the case where seasonality is present in integer-val
ued data and suggest a method of deseasonalizing them. Finally\, an empiri
cal application is carried out.\n\nhttps://indico.uu.se/event/317/contribu
tions/275/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/275/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Fréchet means and Procrustes analysis in Wasserstein space
DTSTART;VALUE=DATE-TIME:20170815T133000Z
DTEND;VALUE=DATE-TIME:20170815T140000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-276@indico.uu.se
DESCRIPTION:Speakers: Yoav Zemel (Ecole polytechnique fédérale de Lausan
ne)\nWe consider three interlinked problems in stochastic geometry: (1) c
onstructing optimal multicouplings of random vectors\; (2) determining th
e Fréchet mean of probability measures in Wasserstein space\; and (3) reg
istering collections of randomly deformed spatial point processes. We dem
onstrate how these problems are canonically interpreted through the prism
of the theory of optimal transportation of measure on $\\mathbb R^d$. We
provide explicit solutions in the one dimensional case\, consistently solv
e the registration problem and establish convergence rates and a (tangent
space) central limit theorem for Cox processes. When $d>1$\, the solution
s are no longer explicit and we propose a steepest descent algorithm for d
educing the Fréchet mean in problem (2). Supplemented by uniform converg
ence results for the optimal maps\, this furnishes a solution to the multi
coupling problem (1). The latter is then utilised\, as in the case $d=1$\
, in order to construct consistent estimators for the registration problem
(3). While the consistency results parallel their one-dimensional counte
rparts\, their derivation requires more sophisticated techniques from conv
ex analysis. This is joint work with Victor M. Panaretos\n\nhttps://indic
o.uu.se/event/317/contributions/276/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/276/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Efficient estimation for diffusions
DTSTART;VALUE=DATE-TIME:20170816T120000Z
DTEND;VALUE=DATE-TIME:20170816T123000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-277@indico.uu.se
DESCRIPTION:Speakers: Nina Munkholt Jakobsen (University of Copenhagen)\nT
his talk concerns estimation of the diffusion parameter of a diffusion pro
cess observed over a fixed time interval. We present conditions on approxi
mate martingale estimating functions under which estimators are consistent
\, rate optimal\, and efficient under high frequency (in-fill) asymptotics
. Here\, limit distributions of the estimators are non-standard in the sen
se that they are generally normal variance-mixture distributions. In parti
cular\, the mixing distribution depends on the full sample path of the dif
fusion process over the observation time interval. Making use of stable co
nvergence in distribution\, we also present the more easily applicable res
ult that estimators normalized by a suitable data-dependent transformation
converge in distribution to a standard normal distribution. The theory is
illustrated by a simulation study.\n\nThe work presented in this talk is
published in:\n\nJakobsen\, N. M. and Sørensen\, M. (2017). *Efficient es
timation for diffusions sampled at high frequency over a fixed time interv
al.* Bernoulli\, 23(3):1874-1910.\n\nhttps://indico.uu.se/event/317/contri
butions/277/
LOCATION:Uppsala University Museum Gustavianum
URL:https://indico.uu.se/event/317/contributions/277/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Best Unbiased Estimators for Doubly Multivariate Data
DTSTART;VALUE=DATE-TIME:20170815T143000Z
DTEND;VALUE=DATE-TIME:20170815T150000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-278@indico.uu.se
DESCRIPTION:Speakers: Arkadiusz Kozioł (Faculty of Mathematics\, Computer
Science and Econometrics University of Zielona Góra\, Szafrana 4a\, 65-5
16 Zielona Góra\, Poland)\nThe article addresses the best unbiased estima
tors of the block compound symmetric covariance\nstructure for m-variate o
bservations with equal mean vector over each level of factor or each time
point (model with structured mean vector). Under multivariate normality\,
the free-coordinate approach is used to obtain unbiased linear and quadrat
ic estimates for the model parameters. Optimality of these estimators foll
ows from sufficiency and completeness of their distributions. Additionally
\, strong consistency is proven. The properties of the estimators in the p
roposed model are compared with the ones in the model with unstructured me
an vector (the mean vector changes over levels of factor or time points).\
n\nhttps://indico.uu.se/event/317/contributions/278/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/278/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Stability of the Spectral EnKF under nested covariance estimators
DTSTART;VALUE=DATE-TIME:20170817T143000Z
DTEND;VALUE=DATE-TIME:20170817T150000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-279@indico.uu.se
DESCRIPTION:Speakers: Marie Turčičová (Charles University\, Prague)\nIn
the case of traditional Ensemble Kalman Filter (EnKF)\, it is known that
the filter error does not \ngrow faster than exponentially for a fixed ens
emble size. The question posted in this contribution is whether the upper
bound for the filter error can be improved by using an improved covariance
estimator that comes from the right parameter subspace and has smaller as
ymptotic variance. Its effect on Spectral EnKF is explored by a simulation
.\n\nhttps://indico.uu.se/event/317/contributions/279/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/279/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Invited Speaker - Random Networks
DTSTART;VALUE=DATE-TIME:20170818T070000Z
DTEND;VALUE=DATE-TIME:20170818T080000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-280@indico.uu.se
DESCRIPTION:Speakers: Svante Janson (Uppsala University)\nhttps://indico.u
u.se/event/317/contributions/280/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/280/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Matrix Independent Component Analysis
DTSTART;VALUE=DATE-TIME:20170814T133000Z
DTEND;VALUE=DATE-TIME:20170814T140000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-281@indico.uu.se
DESCRIPTION:Speakers: Joni Virta (University of Turku)\nIndependent compon
ent analysis (ICA) is a popular means of dimension reduction for vector-va
lued random variables. In this short note we review its extension to arbit
rary tensor-valued random variables by considering the special case of two
dimensions where the tensors are simply matrices.\n\nhttps://indico.uu.se
/event/317/contributions/281/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/281/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Can humans be replaced by computers in taxa recognition?
DTSTART;VALUE=DATE-TIME:20170814T120000Z
DTEND;VALUE=DATE-TIME:20170814T123000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-282@indico.uu.se
DESCRIPTION:Speakers: Johanna Ärje (University of Jyväskylä\, Departmen
t of Mathematics and Statistics)\nBiomonitoring of waterbodies is vital as
the number of anthropogenic stressors on aquatic ecosystems keeps growing
. However\, the continuous decrease in funding makes it impossible to meet
monitoring goals or sustain traditional manual sample processing. We revi
ew what kind of statistical tools can be used to enhance the cost efficien
cy of biomonitoring: We explore automated identification of freshwater mac
roinvertebrates which are used as one indicator group in biomonitoring of
aquatic ecosystems. We present the first classification results of a new i
maging system producing multiple images per specimen. Moreover\, these res
ults are compared with the results of human experts. On a data set of 29 t
axonomical groups\, automated classification produces a higher average acc
uracy than human experts.\n\nhttps://indico.uu.se/event/317/contributions/
282/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/282/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Nonparametric estimation of gradual change points in the jump beha
viour of an Ito semimartingale
DTSTART;VALUE=DATE-TIME:20170814T123000Z
DTEND;VALUE=DATE-TIME:20170814T130000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-283@indico.uu.se
DESCRIPTION:Speakers: Michael Hoffmann (Ruhr-Universität Bochum)\nIn appl
ications the properties of a stochastic feature often change gradually rat
her than\nabruptly\, that is: after a constant phase for some time they sl
owly start to vary. The goal of this talk is to introduce an estimator for
the location of a gradual change point in the jump characteristic of a\nd
iscretely observed Ito semimartingale. To this end we propose a measure of
time variation for the\njump behaviour of the process and consistency of
the desired estimator is a consequence of weak convergence of a suitable e
mpirical process in some function space. Finally\, we discuss simulation r
esults which verify that the new estimator has advantages compared to the
classical argmax-estimator.\n\nhttps://indico.uu.se/event/317/contribution
s/283/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/283/
END:VEVENT
BEGIN:VEVENT
SUMMARY:AIC post-selection inference in linear regression
DTSTART;VALUE=DATE-TIME:20170814T140000Z
DTEND;VALUE=DATE-TIME:20170814T143000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-284@indico.uu.se
DESCRIPTION:Speakers: Ali Charkhi (KULeuven)\nPost-selection inference has
been considered a crucial topic in data\nanalysis. In this article\, we d
evelop a new method to obtain correct inference after model selection by t
he Akaike's information criterion Akaike (1973) in linear regression model
s. Confidence intervals can be calculated by incorporating the randomness
of the model selection in the distribution of the parameter estimators whi
ch act as pivotal quantities. Simulation results show the accuracy of the
proposed method.\n\nhttps://indico.uu.se/event/317/contributions/284/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/284/
END:VEVENT
BEGIN:VEVENT
SUMMARY:E-optimal approximate block designs for treatment-control comparis
ons
DTSTART;VALUE=DATE-TIME:20170817T120000Z
DTEND;VALUE=DATE-TIME:20170817T123000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-298@indico.uu.se
DESCRIPTION:Speakers: Samuel Rosa (Comenius University in Bratislava)\nWe
study $E$-optimal block designs for comparing a set of test treatments wit
h a control treatment. We provide the complete class of all $E$-optimal ap
proximate block designs and we show that these designs are characterized b
y simple linear constraints. Employing the provided characterization\, we
obtain a class of $E$-optimal exact block designs with unequal block sizes
for comparing test treatments with a control.\n\nhttps://indico.uu.se/eve
nt/317/contributions/298/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/298/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Viterbi process for pairwise Markov models
DTSTART;VALUE=DATE-TIME:20170817T080000Z
DTEND;VALUE=DATE-TIME:20170817T083000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-299@indico.uu.se
DESCRIPTION:Speakers: Joonas Sova (University of Tartu)\nMy talk is based
on ongoing joint work with my supervisor Jüri Lember.\n\nWe consider a Ma
rkov chain $Z = \\{Z_k\\}_{k \\geq 1}$ with product\nstate space $\\mathca
l{X}\\times \\mathcal{Y}$\, where $\\mathcal{Y}$ is\na finite set (state s
pace) and $\\mathcal{X}$ is an arbitrary\nseparable metric space (observat
ion space). Thus\, the process $Z$\ndecomposes as $Z=(X\,Y)$\, where $X=\\
{X_k \\}_{k\\geq 1}$ and $Y=\\{Y_k\n\\}_{k\\geq 1}$ are random processes t
aking values in $\\mathcal{X}$\nand $\\mathcal{Y}$\, respectively. Follow
ing\ncite{pairwise\,pairwise2\,pairwise3}\, we call the process $Z$ a\n\\t
extit{pairwise Markov model}. The process $X$ is identified as an\nobserva
tion process and the process $Y$\, sometimes called the \\textit{regime}\,
models the observations-driving hidden state sequence.\nTherefore our ge
neral model contains many well-known stochastic\nmodels as a special case:
hidden Markov models\, Markov\nswitching models\, hidden Markov models wi
th dependent noise and many\nmore. The \\textit{segmentation} or \\textit
{path estimation} problem\nconsists of estimating the realization of $(Y_1
\,\\ldots\,Y_n)$ given a\nrealization $x_{1:n}$ of $(X_1\,\\ldots\,X_n)$.
A standard estimate is\nany path $v_{1:n}\\in \\mathcal{Y}^n$ having maxim
um posterior\nprobability:\n$$v_{1:n}=\\mathop{\\mathrm{argmax}}_{y_{1:n}}
P(Y_{1:n}=y_{1:n}|X_{1:n}=x_{1:n}).$$\nAny such path is called \\textit{V
iterbi path} and we are interested in\nthe behaviour of $v_{1:n}$ as $n$ g
rows. The study of asymptotics of\nViterbi path is complicated by the fact
that adding one more\nobservation\, $x_{n+1}$ can change the whole path\,
and so it is not\nclear\, whether there exists a limiting infinite Viterb
i path. \n\nWe show that under some conditions the infinite Viterbi path i
ndeed exists \nfor almost every realization $x_{1:\\infty}$ of $X$\, there
by defining an infinite Viterbi decoding of $X$\, called the \\textit{Vite
rbi process.} This is done trough construction of \\textit{barriers}. A ba
rrier is a fixed-sized block in the observations $x_{1:n}$ that fixes the
Viterbi path up to\n itself: for every continuation of $x_{1:n}$\, the Vi
terbi path up to\n the barrier remains unchanged. Therefore\, if\nalmost e
very realization of $X$-process contains\ninfinitely many barriers\, then
the Viterbi process exists.\n\nHaving infinitely many barriers is not nece
ssary for\nexistence of infinite Viterbi path\, but the\nbarrier-construct
ion has several advantages. One of them is that it\nallows to construct th
e infinite path \\textit{piecewise}\, meaning that\nto determine the first
$k$ elements $v_{1:k}$ of the infinite path\nit suffices to observe $x_{1
:n}$ for $n$ big enough. Barrier construction has another great advantage:
namely\, the process $(Z\,V)=\\{(Z_k\,V_k)\\}_{k \\geq 1}$\, where $V= \\
{V_k\\}_{k \\geq 1}$ denotes the Viterbi process\, is under certain condit
ions regenerative. This is can be proven by\, roughly speaking\, applying
the Markov splitting method to construct regeneration times for $Z$ which
coincide with the occurrences of barriers. Regenerativity of $(Z\,V)$ allo
ws to easily prove limit theorems to understand the asymptotic behaviour o
f inferences based on Viterbi\npaths. In fact\, in a special case of hidde
n Markov model this regenerative property has already been known to hold a
nd has found several applications cite{AV\,AVacta\,Vsmoothing\,Vrisk\, iow
a}.\n\nhttps://indico.uu.se/event/317/contributions/299/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/299/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Predict extreme influenza epidemics
DTSTART;VALUE=DATE-TIME:20170816T073000Z
DTEND;VALUE=DATE-TIME:20170816T080000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-286@indico.uu.se
DESCRIPTION:Speakers: Maud Thomas (Université Pierre et Marie Curie)\nInf
luenza viruses are responsible for annual epidemics\, causing more than 50
0\,000 deaths per year worldwide. A crucial question for resource planning
in public health is to predict the morbidity burden of extreme epidemics.
We say that an epidemic is extreme whenever the influenza incidence rate
exceeds a high threshold for at least one week. Our objective is to predic
t whether an extreme epidemic will occur in the near future\, say the next
couple of weeks.\n\nThe weekly numbers of influenza-like illness (ILI) in
cidence rates in France are available from the Sentinel network for the pe
riod 1991-2017. ILI incidence rates exhibit two different regimes\, an epi
demic regime during winter and a non-epidemic regime during the rest of th
e year. To identify epidemic periods\, we use a two-state autoregressive h
idden Markov model.\n\nA main goal of Extreme Value Theory is to assess\,
from a series of observations\, the probability of events that are more ex
treme than those previously recorded. Because of the autoregressive struct
ure of the data\, we choose to fit one of the mul- tivariate generalized P
areto distribution models proposed in Rootzén et al. (2016a) [Multivariat
e peaks over threshold models. arXiv:1603.06619v2]\; see also Rootzén et
al. (2016b) [Peaks over thresholds modeling with multivariate generalized
Pareto distributions. arXiv:1612.01773v1]. For these models\, explicit den
sities are given\, and formulas for conditional probabilities can then be
deduced\, from which we can predict if an epidemic will be extreme\, given
the first weeks of observation.\n\nhttps://indico.uu.se/event/317/contrib
utions/286/
LOCATION:Uppsala University Museum Gustavianum
URL:https://indico.uu.se/event/317/contributions/286/
END:VEVENT
BEGIN:VEVENT
SUMMARY:The Elicitation Problem
DTSTART;VALUE=DATE-TIME:20170815T090000Z
DTEND;VALUE=DATE-TIME:20170815T093000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-287@indico.uu.se
DESCRIPTION:Speakers: Tobias Fissler (University of Bern)\nCompeting point
forecasts for functionals such as the mean\, a quantile\, or a certain ri
sk measure are commonly compared in terms of loss functions. These should
be incentive compatible\, i.e.\, the expected score should be minimized by
the correctly specified functional of interest. A functional is called *e
licitable* if it possesses such an incentive compatible loss function. Wit
h the squared loss and the absolute loss\, the mean and the median possess
such incentive compatible loss functions\, which means they are elicitabl
e. In contrast\, variance or Expected Shortfall are not elicitable. \nBesi
des investigating the elicitability of a functional\, it is important to d
etermine the whole class of incentive compatible loss functions as well as
to give recommendations which loss function to use in practice\, taking i
nto regard secondary quality criteria of loss functions such as order-sens
itivity\, convexity\, or homogeneity.\n\nhttps://indico.uu.se/event/317/co
ntributions/287/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/287/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Controlled branching processes in Biology: a model for cell prolif
eration
DTSTART;VALUE=DATE-TIME:20170815T120000Z
DTEND;VALUE=DATE-TIME:20170815T123000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-288@indico.uu.se
DESCRIPTION:Speakers: Carmen Minuesa Abril (University of Extremadura)\nBr
anching processes are relevant models in the development of theoretical ap
proaches to problems in applied fields such as\, for instance\, growth and
extinction of populations\, biology\, epidemiology\, cell proliferation k
inetics\, genetics and algorithm and data structures. The most basic model
\, the so-called Bienaymé-Galton-Watson process\, consists of individuals
that reproduce independently of the others following the same probability
distribution\, known as offspring distribution. A natural generalization
is to incorporate a random control function which determines the number of
progenitors in each generation. The resulting process is called controlle
d branching process.\n\nIn this talk\, we deal with a problem arising in c
ell biology. More specifically\, we focus our attention on experimental da
ta generated by time-lapse video recording of cultured in vitro oligodendr
ocyte cells. In A.Y. Yakovlev et al. (2008) (Branching Processes as Models
of Progenitor Cell Populations and Estimation of the Offspring Distributi
ons\, *Journal of the American Statistical Association*\, 103(484):1357--1
366)\, a two-type age dependent branching process with emigration is consi
dered to describe the kinetics of cell populations. The two types of cell
s considered are referred as type $T_1$ (immediate precursors of oligodend
rocytes) and type $T_2$ (terminally differentiated oligodendrocytes). The
reproduction process of these cells is as follows: when stimulating to div
ide under in vitro conditions\, the progenitor cells are capable of produc
ing either their direct progeny (two daughter cells of the same type) or a
single\, terminally differentiated nondividing oligodendrocyte. Moreover\
, censoring effects as a consequence of the migration of progenitor cells
out of the microscopic field of observation are modelled as the process of
emigration of the type $T_1$ cells.\n\nIn this work\, we propose a two-ty
pe controlled branching process to describe the embedded discrete branchin
g structure of the age-dependent branching process aforementioned. We addr
ess the estimation of the offspring distribution of the cell population in
a Bayesian outlook by making use of disparities. The importance of this p
roblem yields in the fact that the behaviour of these populations is stron
gly related to the main parameters of the offspring distribution and in pr
actice\, these values are unknown and their estimation is necessary. The p
roposed methodology introduced in M. Gonz\\'alez et al. (2017) (Robust est
imation in controlled branching processes: Bayesian estimators via dispari
ties. *Work in progress*)\, is illustrated with an application to the real
data set given in A.Y. Yakovlev et al. (2008).\n\nhttps://indico.uu.se/ev
ent/317/contributions/288/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/288/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Invited Speaker - Independent component analysis using third and f
ourth cumulants
DTSTART;VALUE=DATE-TIME:20170817T110000Z
DTEND;VALUE=DATE-TIME:20170817T120000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-290@indico.uu.se
DESCRIPTION:Speakers: Hannu Oja (University of Turku)\nIn independent comp
onent analysis it is assumed that the observed random variables are linear
combinations of latent\, mutually independent random variables called the
independent components. It is then often thought that only the non-Gaussi
an independent components are of interest and the Gaussian components simp
ly present noise. The idea is then to make inference on the unknown number
of non-Gaussian components and to estimate the transformations back to th
e non-Gaussian components.\n \nIn this talk we show how the classical skew
ness and kurtosis measures\, namely third and fourth cumulants\, can be us
ed in the estimation. First\, univariate cumulants are used as projection
indices \nin search for independent components (projection pursuit\, fastI
CA). Second\, multivariate fourth cumulant matrices are jointly used to so
lve the problem (FOBI\, JADE). The properties of the estimates are conside
red through corresponding optimization problems\, estimating equations\, a
lgorithms and asymptotic statistical properties. The theory is illustrated
with several examples.\n\nhttps://indico.uu.se/event/317/contributions/29
0/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/290/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Simulating and Forecasting Human Population with General Branching
Process
DTSTART;VALUE=DATE-TIME:20170818T093000Z
DTEND;VALUE=DATE-TIME:20170818T100000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-300@indico.uu.se
DESCRIPTION:Speakers: Plamen Trayanov (Sofia Univeristy "St. Kliment Ohrid
ski")\nThe branching process theory is widely used to describe a populatio
n dynamics in which particles live and produce other particles through the
ir life\, according to given stochastic birth and death laws. The theory o
f General Branching Processes (GBP) presents a continuous time model in wh
ich every woman has random life length and gives birth to children in rand
om intervals of time. The flexibility of the GBP makes it very useful for
modelling and forecasting human population. This paper is a continuation o
f previous developments in the theory\, necessary to model the specifics o
f human population\, and presents their application in forecasting the pop
ulation age structure of Bulgaria. It also introduces confidence intervals
of the forecasts\, calculated by GBP simulations\, which reflect both the
stochastic nature of the birth and death laws and the branching process i
tself. The simulations are also used to determine the main sources of risk
to the forecast.\n\nhttps://indico.uu.se/event/317/contributions/300/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/300/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Parameter Estimation for Discretely Observed Infinite-Server Queue
s with Markov-Modulated Input
DTSTART;VALUE=DATE-TIME:20170815T123000Z
DTEND;VALUE=DATE-TIME:20170815T130000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-291@indico.uu.se
DESCRIPTION:Speakers: Birgit Sollie (Vrije Universiteit Amsterdam)\nThe Ma
rkov-modulated infinite-server queue is a queueing system with infinitely
many servers\, where the arrivals follow a Markov-modulated Poisson proces
s (MMPP)\, i.e. a Poisson process with rate modulating between several val
ues. The modulation is driven by an underlying and unobserved continuous t
ime Markov chain $\\{X_t\\}_{t\\geq 0}$. The inhomogeneous rate of the Poi
sson process\, $\\lambda(t)$\, stochastically alternates between $d$ diffe
rent rates\, $\\lambda_1\,\\dots\,\\lambda_d$\, in such a way that $\\lamb
da(t) = \\lambda_i$ if $X_t = i$\, $i=1\,\\dots\,d$.\n\nWe are interested
in estimating the parameters of the arrival process for this queueing syst
em based on observations of the queue length at discrete times only. We as
sume exponentially distributed service times with rate $\\mu$\, where $\\m
u$ is time-independent and known. Estimation of the parameters of the arri
val process has not yet been studied for this particular queueing system.
Two types of missing data are intrinsic to the model\, which complicates t
he estimation problem. First\, the underlying continuous time Markov chain
in the Markov-modulated arrival process is not observed. Second\, the que
ue length is only observed at a finite number of discrete time points. As
a result\, it is not possible to distinguish the number of arrivals and th
e number of departures between two consecutive observations. \n\nIn this t
alk we show how we derive an explicit algorithm to find maximum likelihood
estimates of the parameters of the arrival process\, making use of the EM
algorithm. Our approach extends the one used in Okamura et al. (2009)\, w
here the parameters of an MMPP are estimated based on observations of the
process at discrete times. However\, in contrast to our setting\, Okamura
et al. (2009) do not consider departures and therefore do not deal with t
he second type of missing data. We illustrate the accuracy of the proposed
estimation algorithm with a simulation study.\n\nReference: Okamura H.\,
Dohi T.\, Trivedi K.S.\n(2009).\nMarkovian Arrival Process Parameter Estim
ation With Group Data.\nIEEE/ACM Transactions on Networking. \nVol. 17\, N
o. 4\, pp. 1326--1339\n\nhttps://indico.uu.se/event/317/contributions/291/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/291/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Mallows' Model Based on Lee Distance
DTSTART;VALUE=DATE-TIME:20170817T133000Z
DTEND;VALUE=DATE-TIME:20170817T140000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-292@indico.uu.se
DESCRIPTION:Speakers: Nikolay Nikolov (Institute of Mathematics and Inform
atics\, Bulgarian Academy of Sciences\, Acad. G.Bontchev str.\, block 8\,
1113 Sofia\, Bulgaria)\nIn this paper the Mallows' model based on Lee dist
ance is considered and compared to models induced by other metrics on the
permutation group. As an illustration\, the complete rankings from the Ame
rican Psychological Association election data are analyzed.\n\nhttps://ind
ico.uu.se/event/317/contributions/292/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/292/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Multilevel Functional Principal Component Analysis for Unbalanced
Data
DTSTART;VALUE=DATE-TIME:20170814T143000Z
DTEND;VALUE=DATE-TIME:20170814T150000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-293@indico.uu.se
DESCRIPTION:Speakers: Zuzana Rošťáková (Institute of Measurement Scien
ce\, Slovak Academy of Sciences)\nFunctional principal component analysis
(FPCA) is the key technique for dimensionality reduction and detection of
main directions of variability present in functional data. However\, it is
not the most suitable tool for the situation when analyzed dataset contai
ns repeated or multiple observations\, because information about repeatab
ility of measurements is not taken into account. Multilevel functional pri
ncipal component analysis (MFPCA) is the modified version of FPCA develope
d for data observed at multiple visits. The original MFPCA method was desi
gned for balanced data only\, where for each subject the same number of me
asurements is available. In this article we propose the modified MFPCA alg
orithm which can be applied for unbalanced functional data\; that is\, in
the situation where a different number of observations can be present for
every subject. The modified algorithm is validated and tested on real-worl
d sleep data.\n\nhttps://indico.uu.se/event/317/contributions/293/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/293/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Some recent characterization based goodness of fit tests
DTSTART;VALUE=DATE-TIME:20170815T110000Z
DTEND;VALUE=DATE-TIME:20170815T113000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-294@indico.uu.se
DESCRIPTION:Speakers: Bojana Milošević (Faculty of Mathematics)\nIn this
paper some recent advances in goodness of fit testing are presented. Sp
ecial attention is given to goodness of fit tests based on equidistributio
n and independence characterizations. New concepts are described through s
ome modern exponentiality tests. Their natural generalizations are also pr
oposed. All tests are compared in Bahadur sense.\n\nhttps://indico.uu.se/e
vent/317/contributions/294/
LOCATION: Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/294/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Invited Speaker - Independent component analysis using third and f
ourth cumulants
DTSTART;VALUE=DATE-TIME:20170817T090000Z
DTEND;VALUE=DATE-TIME:20170817T100000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-295@indico.uu.se
DESCRIPTION:Speakers: Hannu Oja (University of Turku)\nIn independent comp
onent analysis it is assumed that the observed random variables are linear
combinations of latent\, mutually independent random variables called the
independent components. It is then often thought that only the non-Gaussi
an independent components are of interest and the Gaussian components simp
ly present noise. The idea is then to make inference on the unknown number
of non-Gaussian components and to estimate the transformations back to th
e non-Gaussian components.\n \nIn this talk we show how the classical skew
ness and kurtosis measures\, namely third and fourth cumulants\, can be us
ed in the estimation. First\, univariate cumulants are used as projection
indices \nin search for independent components (projection pursuit\, fastI
CA). Second\, multivariate fourth cumulant matrices are jointly used to so
lve the problem (FOBI\, JADE). The properties of the estimates are conside
red through corresponding optimization problems\, estimating equations\, a
lgorithms and asymptotic statistical properties. The theory is illustrated
with several examples.\n\nhttps://indico.uu.se/event/317/contributions/29
5/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/295/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Joint Bayesian nonparametric reconstruction of dynamical equations
DTSTART;VALUE=DATE-TIME:20170817T073000Z
DTEND;VALUE=DATE-TIME:20170817T080000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-296@indico.uu.se
DESCRIPTION:Speakers: Christos Merkatas (Department of Mathematics\, Unive
rsity of the Aegean\, Greece)\nWe propose a Bayesian nonparametric mixture
model for the joint full reconstruction of $m$ dynamical equations\, \ngi
ven $m$ observed dynamically-noisy-corrupted chaotic time series. The meth
od of reconstruction is based on the Pairwise Dependent Geometric Stick Br
eaking Processes mixture priors (PDGSBP) first proposed by Hatjispyros et
al. (2017). We assume that\neach set of dynamical equations has a determin
istic part with a known functional form i.e. \n$$\nx_{ji} = g_{j}(\\varthe
ta_j\, x_{j\,i-1}\,\\ldots\,x_{j\,i-l_j}) + \\epsilon_{x_{ji}}\,\\\,\\\,\\
\, 1\\leq j \\leq m\,\\\,\\\, 1\\leq i \\leq n_{j}.\n$$\nunder the assumpt
ion that the noise processes $(\\epsilon_{x_{ji}})$ are independent and id
entically distributed for all $j$ and $i$ from some unknown zero mean proc
ess $f_j(\\cdot)$. Additionally\, we assume that a-priori we have the know
ledge that the processes $(\\epsilon_{x_{ji}})$ for $j=1\,\\ldots\,m$ have
common characteristics\,\ne.g. they may have common variances or even hav
e similar tail behavior etc. For a full reconstruction\, we would like to
jointly estimate the following quantities\n$$\n(\\vartheta_{j})\\in\\Theta
\\subseteq{\\cal R}^{k_j}\,\\quad (x_{j\,0}\,\\ldots\, x_{j\,l_j-1})\\in{\
\cal X}_j\\subseteq{\\cal R}^{l_j}\,\n$$\nand perform density estimation t
o the $m$ noise components $(f_j)$.\n\nOur contention is that whenever the
re is at least one sufficiently large data set\, using carefully selected
informative \nborrowing-of-strength-prior-specifications we are able to r
econstruct those dynamical processes that are responsible for \nthe genera
tion of time series with small sample sizes\; namely sample sizes that are
inadequate for an independent reconstruction.\nWe illustrate the joint es
timation process for the case $m=2$\, when the two time series are coming
from a quadratic and a cubic \nstochastic process of lag one and the noise
processes are zero mean normal mixtures with common components.\n\nhttps:
//indico.uu.se/event/317/contributions/296/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/296/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Confidence regions in Cox proportional hazards model with measurem
ent errors
DTSTART;VALUE=DATE-TIME:20170817T140000Z
DTEND;VALUE=DATE-TIME:20170817T143000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-297@indico.uu.se
DESCRIPTION:Speakers: Oksana Chernova (Taras Shevchenko National Universit
y of Kyiv)\nCox proportional hazards model with measurement errors in cova
riates is considered. It is the ubiquitous technique in biomedical data an
alysis. In Kukush et al. (2011) [ Journal of Statistical Research **45**
\, 77-94 ] and Chimisov and Kukush (2014) [ Modern Stochastics: Theory an
d Applications **1**\, 13-32 ] asymptotic properties of a simultaneous est
imator $(\\lambda_n\;\\beta_n)$ for the baseline hazard rate $\\lambda(\
\cdot)$ and the regression parameter $\\beta$ were studied\, at that the p
arameter set $\\Theta=\\Theta_{\\lambda}\\times \\Theta_{\\beta}$ was assu
med bounded.\n\nIn Kukush and Chernova (2017) [ Theory of Probability and
Mathematical Statistics **96**\, 100-109 ] we dealt with the simultaneous
estimator $(\\lambda_n\;\\beta_n)$ in the case\, where the $\\Theta_{\\l
ambda}$ was unbounded from above and not separated away from $0$. The est
imator was constructed in two steps: first we derived a strongly consisten
t estimator and then modified it to provide its asymptotic normality. \n\n
In this talk\, we construct the confidence interval for an integral funct
ional of $\\lambda(\\cdot)$ and the confidence region for $\\beta$. We r
each our goal in each of the three cases: (a) the measurement error is bo
unded\, (b) it is normally distributed\, or (c) it is a shifted Poisson ra
ndom variable. The censor is assumed to have a continuous pdf. In future r
esearch we intend to elaborate a method for heavy tailed error distributi
ons.\n\nhttps://indico.uu.se/event/317/contributions/297/
LOCATION: Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/297/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Information criteria for structured sparse variable selection
DTSTART;VALUE=DATE-TIME:20170817T123000Z
DTEND;VALUE=DATE-TIME:20170817T130000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-301@indico.uu.se
DESCRIPTION:Speakers: Bastien Marquis (Université Libre de Bruxelles)\nIn
contrast to the low dimensional case\, variable selection under the \nass
umption of sparsity in high dimensional models is strongly influenced by t
he\neffects of false positives.\nThe effects of false positives are temper
ed by combining the variable selection\nwith a shrinkage estimator\, such
as in the lasso\, where the selection is\nrealized by minimizing the sum o
f squared residuals regularized by an $\\ell_1$\nnorm of the selected vari
ables. Optimal variable selection is then equivalent\nto finding the best
balance between closeness of fit and regularity\, i.e.\, to\noptimization
of the regularization parameter with respect to an information\ncriterion
such as Mallows's Cp or AIC. For use in this optimization\nprocedure\, the
lasso regularization is found to be too tolerant towards false\npositives
\, leading to a considerable overestimation of the model size. Using an\n$
\\ell_0$ regularization instead requires careful consideration of the fals
e\npositives\, as they have a major impact on the optimal regularization p
arameter.\nAs the framework of the classical linear model has been analyse
d in previous\nwork\, the current paper concentrates on structured models
and\, more \nspecifically\, on grouped variables. Although the imposed str
ucture in the \nselected models can be understood to somehow reduce the ef
fect of false\npositives\, we observe a qualitatively similar behavior as
in the unstructured\nlinear model.\n\nhttps://indico.uu.se/event/317/contr
ibutions/301/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/301/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Inference on covariance matrices and operators using concentration
inequalities
DTSTART;VALUE=DATE-TIME:20170816T070000Z
DTEND;VALUE=DATE-TIME:20170816T073000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-302@indico.uu.se
DESCRIPTION:Speakers: Adam Kashlak (Cambridge Centre for Analysis\, Univer
sity of Cambridge)\nIn the modern era of high and infinite dimensional dat
a\, classical statistical\nmethodology is often rendered inefficient and i
neffective when confronted\nwith such big data problems as arise in genomi
cs\, medical imaging\, speech\nanalysis\, and many other areas of research
. Many problems manifest when\nthe practitioner is required to take into a
ccount the covariance structure\nof the data during his or her analysis\,
which takes on the form of either a\nhigh dimensional low rank matrix or a
finite dimensional representation of\nan infinite dimensional operator ac
ting on some underlying function space.\nThus\, we propose using tools fro
m the concentration of measure literature\nto construct rigorous descripti
ve and inferential statistical methodology for\ncovariance matrices and op
erators. A variety of concentration inequalities are\nconsidered\, which a
llow for the construction of nonasymptotic dimension-free\nconfidence sets
for the unknown matrices and operators. Given such confidence\nsets a wid
e range of estimation and inferential procedures can be and are\nsubsequen
tly developed.\n\nhttps://indico.uu.se/event/317/contributions/302/
LOCATION:Uppsala University Museum Gustavianum
URL:https://indico.uu.se/event/317/contributions/302/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Methods for bandwidth detection in kernel conditional density esti
mations
DTSTART;VALUE=DATE-TIME:20170815T113000Z
DTEND;VALUE=DATE-TIME:20170815T120000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-303@indico.uu.se
DESCRIPTION:Speakers: Katerina Konecna (Masaryk University)\nThis contribu
tion is focused on the kernel conditional density estimations (KCDE). The
estimation depends on the smoothing parameters which influence the final d
ensity estimation significantly. This is the reason why a requirement of a
ny data-driven method is needed for bandwidth estimation. In this contribu
tion\, the cross-validation method\, the iterative method and the maximum
likelihood approach are conducted for bandwidth selection of the estimator
. An application on a real data set is included and the proposed methods a
re compared.\n\nhttps://indico.uu.se/event/317/contributions/303/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/303/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Invited Speaker - Embedding machine learning in stochastic process
algebra
DTSTART;VALUE=DATE-TIME:20170816T110000Z
DTEND;VALUE=DATE-TIME:20170816T120000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-304@indico.uu.se
DESCRIPTION:Speakers: Jane Hillston (University of Edinburgh)\nhttps://ind
ico.uu.se/event/317/contributions/304/
LOCATION: Museum Gustavianum
URL:https://indico.uu.se/event/317/contributions/304/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Invited speaker - Formal languages for stochastic modelling
DTSTART;VALUE=DATE-TIME:20170816T090000Z
DTEND;VALUE=DATE-TIME:20170816T100000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-305@indico.uu.se
DESCRIPTION:Speakers: Jane Hillston (University of Edinburgh)\nhttps://ind
ico.uu.se/event/317/contributions/305/
LOCATION: Museum Gustavianum
URL:https://indico.uu.se/event/317/contributions/305/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Invited Speaker - Non-limiting spatial extremes
DTSTART;VALUE=DATE-TIME:20170815T070000Z
DTEND;VALUE=DATE-TIME:20170815T080000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-306@indico.uu.se
DESCRIPTION:Speakers: Jenny Wadsworth (Lancaster University)\nMany questio
ns concerning environmental risk can be phrased as spatial extreme value p
roblems. Classical extreme value theory provides limiting models for maxim
a or threshold exceedances of a wide class of underlying spatial processes
. These models can then be fitted to suitably defined extremes of spatial
datasets and used\, for example\, to estimate the probability of events mo
re extreme than we have observed to date. However\, a major practical prob
lem is that frequently the data do not appear to follow these limiting mod
els at observable levels\, and assuming otherwise leads to bias in estimat
ion of rare event probabilities. To deal with this we require models that
allow flexibility in both what the limit should be\, and in the mode of co
nvergence towards it. I will present a construction for such a model and d
iscuss its application to some wave height data from the North Sea.\n\nhtt
ps://indico.uu.se/event/317/contributions/306/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/306/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Invited speaker - Sequential Monte Carlo: basic principles and alg
orithmic inference
DTSTART;VALUE=DATE-TIME:20170814T110000Z
DTEND;VALUE=DATE-TIME:20170814T120000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-307@indico.uu.se
DESCRIPTION:Speakers: Jimmy Olsson (Royal Institute of Technology)\nSequen
tial Monte Carlo methods form a class of genetic-type algorithms sampling\
, on-the-fly and in a very general context\, sequences of probability meas
ures. Today these methods constitute a standard device in the statistician
's tool box and are successfully\n applied within a wide range of scientif
ic and engineering disciplines. This talk is split into two parts\, where
the first provides an introduction to the SMC methodology and the second d
iscusses some novel results concerning the stochastic stability and varian
ce\n estimation in SMC.\n\nhttps://indico.uu.se/event/317/contributions/30
7/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/307/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Invited speaker - Sequential Monte Carlo: basic principles and alg
orithmic inference
DTSTART;VALUE=DATE-TIME:20170814T090000Z
DTEND;VALUE=DATE-TIME:20170814T100000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-308@indico.uu.se
DESCRIPTION:Speakers: Jimmy Olsson (Royal Institute of Technology)\nSequen
tial Monte Carlo methods form a class of genetic-type algorithms sampling\
, on-the-fly and in a very general context\, sequences of probability meas
ures. Today these methods constitute a standard device in the statistician
's tool box and are successfully\n applied within a wide range of scientif
ic and engineering disciplines. This talk is split into two parts\, where
the first provides an introduction to the SMC methodology and the second d
iscusses some novel results concerning the stochastic stability and varian
ce\n estimation in SMC.\n\nhttps://indico.uu.se/event/317/contributions/30
8/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/308/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Delete or Merge Regressors algorithm
DTSTART;VALUE=DATE-TIME:20170815T083000Z
DTEND;VALUE=DATE-TIME:20170815T090000Z
DTSTAMP;VALUE=DATE-TIME:20211130T125002Z
UID:indico-contribution-317-309@indico.uu.se
DESCRIPTION:Speakers: Agnieszka Prochenka (Warsaw University)\nThis paper
addresses a problem of linear and logistic model selection in the presence
of both continuous and categorical predictors. In the literature two type
s of algorithms dealing with this problem can be found. The first one well
known group lasso (\\cite{group}) selects a subset of continuous and a s
ubset of categorical predictors. Hence\, it either deletes or not an entir
e factor. The second one is CAS-ANOVA (\\cite{cas}) which selects a subset
of continuous predictors and partitions of factors. Therefore\, it merges
levels within factors. Both these algorithms are based on the lasso regul
arization.\n\nIn the article a new algorithm called DMR (Delete or Merge R
egressors) is described. Like CAS-ANOVA it selects a subset of continuous
predictors and partitions of factors. However\, instead of using regulariz
ation\, it is based on a stepwise procedure\, where in each step either on
e continuous variable is deleted or two levels of a factor are merged. The
order of accepting consecutive hypotheses is based on sorting t-statistic
s or linear regression and likelihood ratio test statistics for logistic r
egression. The final model is chosen according to information criterion. S
ome of the preliminary results for DMR are described in \\cite{pro}.\n\nDM
R algorithm works only for data sets where $p \n\nhttps://indico.uu.se/eve
nt/317/contributions/309/
LOCATION:Uppsala University Ångströmslaboratoriet
URL:https://indico.uu.se/event/317/contributions/309/
END:VEVENT
END:VCALENDAR