Monday Lunch Meetings for the section SPT
The meetings are on Mondays about every fourth week at 12:15 in Aud. 10.
Previous
talks: 2023-20242022-2023 2021-2022 2020-2021 2019-2020 2018-2019 2017-2018; 2016-2017; 2015-2016; 2014-2015; 2013-2014;
2012-2013; 2011-2012;
2010-2011; 2009-2010
3/2/2025: Susanne Ditlevsen:
13/1/2025: NN:
16/12/2024: Cecilie Olesen Recke: Identifiability and Estimation in Lyapunov Models
25/11/2024: Sebastian Weichwald: ...
28/10/2024: Alex Markham: Identifiability of causal factor models
30/9/2024: Michael Sørensen: Toroidal diffusions with a view to applications in biology
Previous talks, 2023-2024
18/12/2023: Niklas Pfister: Extrapolation: What happens outside of the support?
20/11/2023: Jeff Adams: Adjusting for Multi-Cause Confounding
23/10/2023: Matthieu Bulté: An Autoregressive Model for Time Series of Random Objects
Previous talks, 2022-2023
12/6/2023: Shimeng Huang: Causal Change Point Detection and Localization
22/5/2023: Lucas Kook: Invariant causal prediction for non-additive noise models
17/4/2023: Jonas Gyde Hermansen: Stochastic Differential Equations with Random Effects: Existence and Uniqueness
13/3/2023: Alexander Christgau: Efficient adjustment for unstructured covariates
20/2/2023: Anton Rask Lundborg: Modern Methods for Variable Significance Testing
9/1/2023: Myrto Limnios: A Semiparametric Model for Testing Conditional Local Independence of Point Processes
19/12/2022: Niels Richard Hansen: Graphical Tensor Equations (with gløgg)
24/10/2022: Christian Holberg: Asymptotics of Cointegration Estimator Under Different Rank Specifications
26/9/2022: Leonard Henckel: Estimating causal effects under interference and confounding
29/8/2022: Nikolaj Thams:
Evaluating Robustness to Dataset Shift via Parametric Robustness Sets
Previous talks, 2021-2022
20/6/2022: Predrag Pilipovic: Splitting Schemes for Ergodic One-sided Lipschitz Diffusion: Numerical Properties and Parameter Estimation
23/5/2022: Marie Leváková: Cointegration as a novel approach to analyze EEG signals
25/4/2022: Susanne Ditlevsen: Time scales in early warnings: a probabilistic approach
Previous talks, 2020-2021
30/11/2020: Robyn Stuart: Controlling COVID-19 via test-trace-quarantine, part II
16/11/2020: Cuong Ngo: Buzz detection using accelerometer data with supervised neural networks
2/11/2020: Robyn Stuart: Controlling COVID-19 via test-trace-quarantine
19/10/2020: Jan van Waaij: Uncertainty quantification for community detection
5/10/2020: Steffen Lauritzen: Total positivity in exponential families with application to structured binary distributions.
21/9/2020: Susanne Ditlevsen: What can we learn from stochastic modeling of neurotransmitter release?
7/9/2020: Helle Sørensen: Time series with discrete outcomes
Previous talks, 2019-2020
25/5/2020: Chuang Xu: One-Dimensional Continuous Time Markov Chains with Unbounded Jumps
11/5/2020: Linard Hoessly: Stationary distributions of stochastic reaction networks via decomposition
27/4/2020: Beatriz Pascual Escudero: Using Algebraic Geometry: Concentration Robustness in Reaction Networks
30/3/2020: Anders Tolver: Corona-cancelled
17/2/2020: Rune Christiansen: Towards causal inference for spatio-temporal data: adjusting for time-invariant latent confounders
3/2/2020: Gherardo Varando: The R Package stagedtrees for Structural Learning of Stratified Staged Trees
20/1/2020: Phillip Bredahl Mogensen: Causal discovery and residual entropy estimation
9/12/2019: C4C
Competition Team: What was the competition about, and what did we do?
25/11/2019: Marie Leváková: Neuronal coding, Fisher information and “good” noise
11/11/2019: Carsten Wiuf: Molecular machines and the EM algorithm
28/10/2019: Sebastian Weichwald: Confounding-Robust ICA — Adjusting for Group-Wise Stationary Noise
30/9/2019: Bo Markussen: Analysis of faeces particles size distribution from ruminating species using linear modelling in the Aitchison geometry
16/9/2019: Steffen Lauritzen: Harmonic Analysis of Random Graphs
Previous talks, 2018-2019
24/6/2019: Mathias Drton: Testing independence
20/5/2019: Gherardo Varando: Estimation of Dynamical Systems from Cross Sectional Data
29/4/2019: Robyn Stuart: Software for compartmental models
1/4/2019: Lasse Petersen: Towards Non-Parametric Causal
Discovery
18/3/2019: Chuang Xu: One-dimensional stochastic reaction networks:
Classification and dynamics
Stochastic reaction networks are widely used to model various
biochemical phenomena. To understand their long-term stochastic
dynamics, stationary distributions are often computed. One crucial
dynamical property to guarantee the existence of a stationary
distribution is positive recurrence. However, it is not easy to
provide checkable criteria for stochastic reaction networks, only by
topological or graphical structures.
Motivated by this need, this talk contributes to stochastic dynamics
of chemical reaction networks (CRNs) with one-dimensional
stoichiometric subspace. I will first present a classification of the
state space of the underlying continuous time Markov chain (CTMC) and
mention how to use this result to discuss the diversity of long-term
dynamics of stochastic CRNs.
Moreover, I will present checkable necessary and sufficient network
conditions for various dynamical properties: Recurrence (positive and
null), transience, (non)explosivity, (non)implosivity, as well as
existence of moments of passage times. As a byproduct, any
one-dimensional weakly reversible CRN is positive recurrent,
confirming the Positive Recurrence Conjecture proposed by Anderson and
Kim in 2018 (in 1-d case).
Finally, I will emphasize results on one-species CRNs, regarding
stationary distributions and present parameter regions for consistency
and inconsistency of stochastic and deterministic one-species CRNs
regarding various dynamical properties aforementioned.
18/2/2019: Line Kühnel: Stochastic Modelling on Manifolds
4/2/2019: Angelica Torres Bustos: Parameter regions for bistability in Chemical Reaction Networks.
21/1/2019: Susanne Ditlevsen: ...
10/12/2018: Jimmy Olsson: Partial ordering of inhomogeneous
Markov chains with applications to Markov chain Monte Carlo
methods
This talk discusses the asymptotic variance of sample path averages for inhomogeneous Markov chains that evolve alternatingly according to two different π-reversible Markov transition kernels P and Q. More specifically, our main result allows us to compare directly the asymptotic variances of two inhomogeneous Markov chains associated with different kernels P_i and Q_i , i ∈ {0, 1}, as soon as the kernels of each pair (P_0, P_1) and (Q_0, Q_1) can be ordered in the sense of lag-one autocovariance. As an important application, we use this result for comparing different data-augmentation-type Metropolis-Hastings algorithms. In particular, we compare some pseudo-marginal algorithms and propose a novel exact algorithm, referred to as the random shake algorithm, which is more efficient, in terms of asymptotic variance, than the Grouped Independence Metropolis-Hastings algorithm and has a computational complexity that does not exceed that of the Monte Carlo within Metropolis algorithm.
26/11/2018: Laura Battagliola: Quantile Regression for Longitudinal Functional Data
12/11/2018: Niels Richard Hansen: Learning large scale ordinary
differential equation systems
29/10/2018: Elisenda Feliu: Signs of polynomials and
dynamical properties of reaction networks
12.50: IT Learning Center.
15/10/2018: Cancelled, vacation
01/10/2018: Kang Li: The talk is about Inference for
complicated mixtures
Mixture models arise when we assume the observation is driven by a discrete
number of hidden components. Here we consider inference of mixtures where
the observation distributions conditional on hidden components are complicated,
and the log-likelihood is computationally expensive and highly non-convex. Given
the observation model (with
known parameters) and observed data, we aim to cluster the data according to
the hidden components, as well as to estimate the parameters describing each
component. Standard methods with maximum log-marginal-likelihood
and the expectation-maximization (EM) algorithm using random starting parameters,
perform poorly due to the expensive and non-convex objective function. In this study
we attempt to overcome the difficulties in complicated mixture models. We
recommend the hard-assigned EM for complicated mixtures to save computational
burden, and propose two parameter initialization schemes: one extending the
$k$-means++ to consider arbitrary model distribution, the other relying on
data pre-clustering in a space of log-likelihood distances that describe relationships
among data. Simulation studies
are conducted in a neuroscience and visual attention environment considering
three distinct model types with different optimization methods, and results show the
proposed methods provide consistently better performance in all studies.
Previous talks, 2017-2018
04/09/2017: Jacob Østergaard: Capturing spike variability in noisy Izhikevich neurons using point process Generalized Linear Models
18/09/2017: Luis Ernesto Salasar: A nonparametric Bayesian approach for the two-sample problem
02/10/2017: Samuele Soraggi: Detection of ploidy levels from next generation sequencing data
16/10/2017: Cancelled, autumn vacation
30/10/2017: Søren Wengel Mogensen: Learning causal structure in dynamical systems
13/11/2017: Bo Markussen: A multiple testing procedure
27/11/2017: Rune Christiansen: Invariant Causal Prediction with Latent Variables
11/12/2017: Frederik Vissing Mikkelsen: Extending SURE to Estimators with Data Adaptive Model Selection via Flows
08/01/2018: Bob Pepin: A Non-Stationary Ergodic Theorem
22/01/2018: Niels Aske Lundtorp Olsen: False Discovery Rates for Functional Data
05/02/2018: Niels Richard Hansen: On the size of introns in
vertebrates. Or how to fit additive models with a large number of categories
19/02/2018: Marcin Mider: Computational Aspects of the Unbiased Inference for Discretely Observed Diffusions
05/03/2018: Amirhossein Sadeghi Manesh: Expected number of solutions to a system of polynomial equations with random coefficients
19/03/2018: Michael Sørensen: Toroidal diffusions with a view to protein structure
02/04/2018: Cancelled, Easter
16/04/2018: Cancelled, Evaluation of Institute
30/04/2018: Margaret Taub: Telomere length estimation and analysis from WGS data
14/05/2018: Therese Graversen: An expectation calculator based on probability propagation
28/05/2018: Cancelled.
11/06/2018: Steffen Lauritzen: Max-linear Bayesian networks
25/06/2018: Summer vacation
Previous talks, 2016-2017
05/09/2016: Abhishek Pal Majumder
19/09/2016: Mareile Große Ruse
03/10/2016: Anders Tolver
17/10/2016: Cancelled, autumn vacation
31/10/2016: Nina Munkholt: Efficient estimation for diffusions with jumps
14/11/2016: Bob Pepin: A Quantitative Averaging Principle for Temperature-Accelerated Molecular Dynamics
28/11/2016: Steffen Lauritzen: Thiele's Convent Insurance:
Assessing hazard of marriage
12/12/2016: Niels Richard Hansen: Why AIC doesn't do what you think and what can be done about it
9/01/2017: Janne Kool: Activation and repression in gene-regulation
23/01/2017: Mareile Große Ruse: Maximum-likelihood estimation
for multi-dimensional inhomogeneous stochastic differential equations
with mixed effects
06/02/2017: Johannes Heiny: Estimation of extreme eigenvalues of sample covariance and correlation matrices
20/02/2017: Jonas Peters: Kernel-based Tests for Joint Independence
06/03/2017: Abhishek Pal Majumder: Connection of Stochastic
recurrence equation to Reaction network and some its properties
20/03/2017: Helle Sørensen: Interval-wise testing for functional data
03/04/2017: Emil S. Jørgensen: Prediction-Based Estimation of Ergodic Diffusions with High-Frequency Data
17/04/2017: Cancelled, Easter
01/05/2017: Cancelled, Holiday
15/05/2017: Meritxell Saez: Positive linear elimination in chemical reaction networks
29/05/2017: Amirhossein Sadeghi Manesh: Intermediates, Binomiality and Multistationarity
12/06/2017: Mads Bonde Raad
26/06/2017: Summer vacation
Previous talks, 2015-2016
14/09/2015: Susanne Ditlevsen: Multi-class oscillating systems of interacting neurons
28/09/2015: Michael Marcondes de Freitas
12/10/2015: Cancelled, autumn vacations
26/10/2015: Meritxell Saez Cornellana
9/11/2015: Bob Pepin
23/11/2015: Samuele Soraggi
07/12/2015: Jacob Østergaard
11/01/2016: Mareile Große Ruse
25/01/2016: Catalina Vich Llompart
08/02/2016: Eduardo Garcia Portugues
22/02/2016: Olivier Wintenberger
07/03/2016: Kang Li
21/03/2016: Cancelled, Easter
04/04/2016: Bo Markussen
18/04/2016: Pierre Julien Gaillard
02/05/2016: Frederik Riis Mikkelsen
16/05/2016: Cancelled, Holiday
30/05/2016: Niels Aske Lundtorp Olsen
13/06/2016: Robyn Stuart
27/06/2016: Summer vacation
Previous talks, 2014-2015
15/09/2014: Niels Richard Hansen: Degrees of freedom for nonlinear least squares estimation
29/09/2014: Kang Li
13/10/2014: Cancelled, autumn vacations
27/10/2014: Lars Lau Raket
10/11/2014: Jesper Lund Pedersen
24/11/2014: Rolf Poulsen: Volatility is log-normal but not for the
reason you think
08/12/2014: Helle Sørensen
05/01/2015: Bo Markussen: operatorCalc - an R package for integral operators on multivariate function space
19/01/2015: Jiawen Gu
02/02/2015: Johannes Heiny
16/02/2015: Robyn Stuart
02/03/2015: Benjamin Guedj
16/03/2015: Steffen Lauritzen: Graphical models for random exchangeable networks
30/03/2015: Cancelled, Easter
13/04/2015: Cancelled due to PhD course
27/04/2015: Olivier Wintenberger: Robustness and optimality for prediction with expert advice
11/05/2015: Cancelled due to meeting with Science Advisory Board
25/05/2015: Cancelled, Holiday
08/06/2015: Frederik Riis Mikkelsen
22/06/2015: Summer vacation
06/07/2015: Summer vacation
Previous talks, 2013-2014
16/09/2013: Carsten Wiuf
30/09/2013: Trine Krogh Boomsma
14/10/2013: Cancelled, autumn vacations
28/10/2013: Jeffrey Collamore
11/11/2013: Massimiliano Tamborrino
25/11/2013: Kristian Buchardt
09/12/2013: Salvador Pineda
06/01/2014: Adam Lund
20/01/2014: Andre Ribeiro: How can we price options on realized variance?
03/02/2014: Alexandre Iolov
17/02/2014: Michael Sørensen
03/03/2014: Nina Munkholt
17/03/2014: Ninna Reitzel Jensen
31/03/2014: Daniele Cappelletti
14/04/2014: Cancelled, Easter
28/04/2014: Mousavi Seyed Nourollah
12/05/2014: Rolf Poulsen
26/05/2014: Martin Vincent
09/06/2014: Cancelled, Holiday
23/06/2014: Summer vacation
07/07/2014: Summer vacation
Previous talks, 2012-2013
03/09/2012: Flemming Topsøe: An algorithm for isotonic
regression in trees with an application to a problem of universal
coding
17/09/2012: Susanne Ditlevsen: Parameter estimation in the
stochastic Morris-Lecar neuronal model with particle filter
methods. Slides
01/10/2012: Jesper Lund Pedersen
15/10/2012: Cancelled, autumn vacations
29/10/2012: Yuwei Zhao
12/11/2012: Miguel Angel Alejo Plana: On networks of interacting species
26/11/2012: Alexandre Iolov: Optimal control of inter-spike intervals
given minimal
feedback. Slides
10/12/2012: Michael Sørensen
07/01/2013: Mogens Steffensen
21/01/2013: Elisenda Feliu
04/02/2013: Rolf Poulsen
18/02/2013: Alexander Sokol
04/03/2013: Niels Richard Hansen
18/03/2013: Sima Mashayeki
01/04/2013: Cancelled, Easter
15/04/2013: Anders Tolver
29/04/2013: Jostein Paulsen
13/05/2013: Helle Sørensen
27/05/2013: Anders Rønn-Nielsen
10/06/2013: Bo Markussen: Functional regression without penalization
24/06/2013: Summer vacation
Previous talks, 2011-2012
05/09/2011: Morten Tolver Kronborg
19/09/2011: Helle Sørensen
03/10/2011: Martin Hunting
17/10/2011: Cancelled, autumn vacations
31/10/2011: Anders Christian Jensen
14/11/2011: Michael Sørensen
28/11/2011: Kamille Sofie Tågholt
12/12/2011: Mogens Steffensen
09/01/2012: Stefan Mihalache
23/01/2012: Jeffrey Collamore
06/02/2012: Rolf Poulsen: Approximation Behooves Calibration
20/02/2012: Morten Karlsmark
05/03/2012: Niels Richard Hansen
19/03/2012: Susanne Ditlevsen
02/04/2012: Cancelled, Easter
16/04/2012: Trine Krogh Boomsma
30/04/2012: Carsten Wiuf
14/05/2012: Sam Finch
28/05/2012: Cancelled, holiday
11/06/2012: Cancelled, Nordstat meeting
25/06/2012: Summer vacation
Previous talks, 2010-2011
06/09/2010: Flemming Topsøe: Cognition and Inference
20/09/2010: Randi Grøn
04/10/2010: Niels Richard Hansen
18/10/2010: Massimiliano Tamborrino
01/11/2010: Thomas Mikosch: The Extremogram: A Corellogram for Extreme Events
With Richard A. Davis (Columbia). Abstract: The extremogram measures serial tail dependence in a
time series. It has the interpretation as a limiting correlogram derived from
conditional probabilities. This definition opens the door to
classical time series analysis, including the spectral analysis
of extreme events. We also discuss the estimation of the correlogram by a
sample analog. The stationary bootstrap of Politis and Romano (JASA 1994) is a
useful technique for constructing confidence bands for the sample extremogram.
15/11/2010: Cindy Greenwood
29/11/2010: Jessica Kasza
13/12/2010: Martin Jacobsen: Stationaritet af ARCH(m)-processer
10/01/2011: Anders Christian Jensen
24/01/2011: Rolf Poulsen: Optimal Portfolio Choice Under Partial Information and Transaction Costs. Or: How to Invest When You Don't Know What to Expect
07/02/2011: Jostein Paulsen: On the optimal dividend problem for diffusion processes
21/02/2011: Anders Rønn-Nielsen
07/03/2011: CANCELLED. Alexander Sokol: Novikov's betingelse for martingaler
med spring
21/03/2011: Mogens Steffensen: ACTULUS: An introduction to our
project with Edlund, the IT University,
and the Danish National Advanced Technology Foundation.
04/04/2011: Martin Vincent
18/04/2011: Cancelled, Easter vacations
02/05/2011: Alexander Sokol: Novikov's betingelse for martingaler
med spring
16/05/2011: Kenneth Bruhn: A Consumption and Investment Problem
with Recursive Utility and
Multiplicative Habit Formation of Past Utility
30/05/2011: Jessica Kasza: Dependence of gene sets in gastric
cancers
13/06/2011: Cancelled, Pinse holiday
Summer vacations
Previous talks, 2009-2010
28/09/2009: Louise Kallehauge
12/10/2009: Susanne Ditlevsen
26/10/2009: Michael Sørensen
09/11/2009: Cathrine Jessen
23/11/2009: Cancelled because there are 3 ordinary seminars very close in time
07/12/2009: Patrick Jahn
Christmas vacations
11/01/2010: Mogens Steffensen
25/01/2010: Niels Richard Hansen
08/02/2010: Esben Kryger
22/02/2010: Jessica Kasza
08/03/2010: Rolf Poulsen
22/03/2010: Jesper Lund Pedersen
05/04/2010: Cancelled, Easter vacations
19/04/2010: Jeffrey Collamore
03/05/2010: Helle Sørensen
17/05/2010: Martin Jacobsen: Boundary Crossings for some Integrated Processes
31/05/2010: Lisbeth Carstensen
14/06/2010: Cancelled, Nordstat meeting
Summer vacations
Back to homepage of Susanne
Ditlevsen