In vielen praktischen Anwendungen können die interessierenden Grössen nicht unmittelbar beobachtet werden, oder sie müssen von anderen Grössen abgeleitet werden. Oftmals sind die Beobachtungen rauschbehaftet, oder die Rekonstruktion die interessierenen Grössen ist instabil. In diesen Fällen spricht man von schlechtgestellten (inversen) Problemen. Derartige Probleme treten insbesondere in der Computertomographie oder anderen bildverarbeitenden Verfahren auf, aber die Kreis der Anwendungen ist viel breiter. Besonders erwähnenswert ist die Modelkalibrierung. Hier müssen betrachtete Modelle, die von weiteren Parametern abhängen an vorliegende Daten angepasst werden.
Bei der Rekonstruktion von interessierenden Grössen aus (statistisch) verrauschten Daten handelt es sich um ein statistisches inverses Problem. In diesem Falle handelt es sich oft um ein (nichtparametrisches) Schätzproblem, das in traditioneller Weise gelöst werden kann, welches aber auch durch eine Bayessche Analyse betrachtet werden kann. Ein derartiges Herangehen liefert neben dem eigentlichen Schätzer noch weitere Informationen, die zur Quantifizierung der dem Problem inherenten Unsicherheit genutzt werden können.
Am Weierstrass Institut reichen die Untersuchungen in diesem Themenkomplex von grundlegenden Forschungen bis hin zu problemspezifischen Fragestellungen. Berücksichtigt werden hier Kenntnisse aus der Regularisierungstheorie, der nichtparametrischen Statistik, und aus den jeweiligen spezifischen Anwendungen.
Seit mehreren Jahren liegt der Schwerpunkt auf der Untersuchung sowohl linearer als auch nichtlinearer statistischer inverser Probleme. Gesucht werden effiziente Rekonstruktionsverfahren. Viele Rekonstruktionsverfahren (Regularisierungen) nutzen einen spezifischen Regularisierungsparameter, und die korrekte Bestimmung des Parameters (Modelselektion) ist wichtig. Weiterhin werden Bayessche Verfahren, sowohl theoretisch als auch in spezifischen Anwendungen (MCMCVerfahren) untersucht. Eine wichtige Anwendung ist im Rahmen des Industrieprojekts BOP.
Publikationen
Monografien

M. Hintermüller, K. Papafitsoros, Chapter 11: Generating Structured Nonsmooth Priors and Associated Primaldual Methods, in: Processing, Analyzing and Learning of Images, Shapes, and Forms: Part 2, R. Kimmel, X.Ch. Tai, eds., 20 of Handbook of Numerical Analysis, Elsevier, 2019, pp. 437502, (Chapter Published), DOI 10.1016/bs.hna.2019.08.001 .

M. Hintermüller, M. Hinze, J. Sokołowski, S. Ulbrich, eds., Special issue to honour Guenter Leugering on his 65th birthday, 1 of Control & Cybernetics, Systems Research Institute, Polish Academy of Sciences, Warsaw, 2019, (Collection Published).

M. Hintermüller, J.F. Rodrigues, eds., Topics in Applied Analysis and Optimisation  Partial Differential Equations, Stochastic and Numerical Analysis, CIM Series in Mathematical Sciences, Springer Nature Switzerland AG, Cham, 2019, 396 pages, (Collection Published).

P. Deuflhard, M. Grötschel, D. Hömberg, U. Horst, J. Kramer, V. Mehrmann, K. Polthier, F. Schmidt, Ch. Schütte, M. Skutella, J. Sprekels, eds., MATHEON  Mathematics for Key Technologies, 1 of EMS Series in Industrial and Applied Mathematics, European Mathematical Society Publishing House, Zurich, 2014, 453 pages, (Collection Published).
Artikel in Referierten Journalen

P. Mathé, Bayesian inverse problems with noncommuting operators, Mathematics of Computation, 88 (2019), pp. 28972912, DOI 10.1090/mcom/3439 .
Abstract
The Bayesian approach to illposed operator equations in Hilbert space recently gained attraction. In this context, and when the prior distribution is Gaussian, then two operators play a significant role, the one which governs the operator equation, and the one which describes the prior covariance. Typically it is assumed that these operators commute. Here we extend this analysis to noncommuting operators, replacing the commutativity assumption by a link condition. We discuss its relation to the commuting case, and we indicate that this allows to use interpolation type results to obtain tight bounds for the contraction of the posterior Gaussian distribution towards the data generating element. 
M. Eigel, M. Marschall, R. Schneider, Samplingfree Bayesian inversion with adaptive hierarchical tensor representations, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 34 (2018), pp. 035010/1035010/29, DOI 10.1088/13616420/aaa998 .
Abstract
The statistical Bayesian approach is a natural setting to resolve the illposedness of inverse problems by assigning probability densities to the considered calibration parameters. Based on a parametric deterministic representation of the forward model, a samplingfree approach to Bayesian inversion with an explicit representation of the parameter densities is developed. The approximation of the involved randomness inevitably leads to several high dimensional expressions, which are often tackled with classical sampling methods such as MCMC. To speed up these methods, the use of a surrogate model is beneficial since it allows for faster evaluation with respect to calibration parameters. However, the inherently slow convergence can not be remedied by this. As an alternative, a complete functional treatment of the inverse problem is feasible as demonstrated in this work, with functional representations of the parametric forward solution as well as the probability densities of the calibration parameters, determined by Bayesian inversion. The proposed samplingfree approach is discussed in the context of hierarchical tensor representations, which are employed for the adaptive evaluation of a random PDE (the forward problem) in generalized chaos polynomials and the subsequent highdimensional quadrature of the loglikelihood. This modern compression technique alleviates the curse of dimensionality by hierarchical subspace approximations of the involved low rank (solution) manifolds. All required computations can be carried out efficiently in the lowrank format. A priori convergence is examined, considering all approximations that occur in the method. Numerical experiments demonstrate the performance and verify the theoretical results. 
A. Anikin, A. Gasnikov, P. Dvurechensky, A. Turin, A. Chernov, Dual approaches to the minimization of strongly convex functionals with a simple structure under affine constraints, Computational Mathematics and Mathematical Physics, 57 (2017), pp. 12621276.

D. Belomestny, H. Mai, J.G.M. Schoenmakers, Generalized PostWidder inversion formula with application to statistics, Journal of Mathematical Analysis and Applications, 455 (2017), pp. 89104.
Abstract
In this work we derive an inversion formula for the Laplace transform of a density observed on a curve in the complex domain, which generalizes the well known PostWidder formula. We establish convergence of our inversion method and derive the corresponding convergence rates for the case of a Laplace transform of a smooth density. As an application we consider the problem of statistical inference for variancemean mixture models. We construct a nonparametric estimator for the mixing density based on the generalized PostWidder formula, derive bounds for its root mean square error and give a brief numerical example. 
S. Bürger, P. Mathé, Discretized Lavrent'ev regularization for the autoconvolution equation, Applicable Analysis. An International Journal, 96 (2017), pp. 16181637, DOI 10.1080/00036811.2016.1212336 .
Abstract
Lavrent?ev regularization for the autoconvolution equation was considered by Janno J. in Lavrent?ev regularization of illposed problems containing nonlinear neartomonotone operators with application to autoconvolution equation, Inverse Prob. 2000;16:333?348. Here this study is extended by considering discretization of the Lavrent?ev scheme by splines. It is shown how to maintain the known convergence rate by an appropriate choice of spline spaces and a proper choice of the discretization level. For piecewise constant splines the discretized equation allows for an explicit solver, in contrast to using higher order splines. This is used to design a fast implementation by means of postsmoothing, which provides results, which are indistinguishable from results obtained by direct discretization using cubic splines. 
M. Hintermüller, C.N. Rautenberg, T. Wu, A. Langer, Optimal selection of the regularization function in a generalized total variation model. Part II: Algorithm, its analysis and numerical tests, Journal of Mathematical Imaging and Vision, 59 (2017), pp. 515533.
Abstract
Based on the generalized total variation model and its analysis pursued in part I (WIAS Preprint no. 2235), in this paper a continuous, i.e., infinite dimensional, projected gradient algorithm and its convergence analysis are presented. The method computes a stationary point of a regularized bilevel optimization problem for simultaneously recovering the image as well as determining a spatially distributed regularization weight. Further, its numerical realization is discussed and results obtained for image denoising and deblurring as well as Fourier and wavelet inpainting are reported on. 
M. Hintermüller, C.N. Rautenberg, Optimal selection of the regularization function in a weighted total variation model. Part I: Modeling and theory, Journal of Mathematical Imaging and Vision, 59 (2017), pp. 498514.
Abstract
Based on the generalized total variation model and its analysis pursued in part I (WIAS Preprint no. 2235), in this paper a continuous, i.e., infinite dimensional, projected gradient algorithm and its convergence analysis are presented. The method computes a stationary point of a regularized bilevel optimization problem for simultaneously recovering the image as well as determining a spatially distributed regularization weight. Further, its numerical realization is discussed and results obtained for image denoising and deblurring as well as Fourier and wavelet inpainting are reported on. 
P. Mathé, S.V. Pereverzev, Complexity of linear illposed problems in Hilbert space, Journal of Complexity, 38 (2017), pp. 5067.

D. Belomestny, J.G.M. Schoenmakers, Statistical inference for timechanged Lévy processes via Mellin transform approach, Stochastic Processes and their Applications, 126 (2016), pp. 20922122.

K. Lin, S. Lu, P. Mathé, Oracletype posterior contraction rates in Bayesian inverse problems, Inverse Problems and Imaging, 9 (2015), pp. 895915.

P. Mathé, Adaptive discretization for signal detection in statistical inverse problems, Applicable Analysis. An International Journal, 94 (2015), pp. 494505.

S. Anzengruber, B. Hofmann, P. Mathé, Regularization properties of the sequential discrepancy principle for Tikhonov regularization in Banach spaces, Applicable Analysis. An International Journal, 93 (2014), pp. 13821400.

S. Lu, P. Mathé, Discrepancy based model selection in statistical inverse problems, Journal of Complexity, 30 (2014), pp. 290308.

C. Marteau, P. Mathé, General regularization schemes for signal detection in inverse problems, Mathematical Methods of Statistics, 23 (2014), pp. 176200.

R.I. Boţ, B. Hofmann, P. Mathé, Regularizability of illposed problems and the modulus of continuity, Analysis and Applications, 32 (2013), pp. 299312.

E. Burnaev, A. Zaytsev, V. Spokoiny, Nonasymptotic properties for Gaussian field regression, Automation and Remote Control, 74 (2013), pp. 16451655.

Q. Jin, P. Mathé, Oracle inequality for a statistical RausGfrerertype rule, SIAM ASA J. Uncertainty Quantification, 1 (2013), pp. 386407.

A. Zaitsev, E. Burnaev, V. Spokoiny, Properties of the posterior distribution of a regression model based on Gaussian random fields, Automation and Remote Control, 74 (2013), pp. 16451655.

B. Hofmann, P. Mathé, Some note on the modulus of continuity for illposed problems in Hilbert space, Trudy Instituta Matematiki i Mekhaniki UrO RAN, 18 (2012), pp. 3441.

G. Blanchard, P. Mathé, Discrepancy principle for statistical inverse problems with application to conjugate gradient iteration, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 28 (2012), pp. 115011/1115011/23.

B. Hofmann, P. Mathé, Parameter choice in Banach space regularization under variational inequalities, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 28 (2012), pp. 104006/1104006/17.

S.M.A. Becker, Regularization of statistical inverse problems and the Bakushinskii veto, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 27 (2011), pp. 115010/1115010/22.
Abstract
In the deterministic context Bakushinskii's theorem excludes the existence of purely datadriven convergent regularization for illposed problems. We will prove in this work that in the statistical setting we can either construct a counter example or develop an equivalent formulation depending on the considered class of probability distributions. Hence, Bakushinskii's theorem does not generalize to the statistical context, although this has often been assumed in the past. To arrive at this conclusion, we will deduce from the classic theory new concepts for a general study of statistical inverse problems and perform a systematic clarification of the key ideas of statistical regularization. 
F. Bauer, P. Mathé, Parameter choice methods using minimization schemes, Journal of Complexity, 27 (2011), pp. 6885.
Abstract
In this paper we establish a generalized framework, which allows to prove convergenence and optimality of parameter choice schemes for inverse problems based on minimization in a generic way. We show that the well known quasioptimality criterion falls in this class. Furthermore we present a new parameter choice method and prove its convergence by using this newly established tool. 
J. Flemming, B. Hofmann, P. Mathé, Sharp converse results for the regularization error using distance functions, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 27 (2011), pp. 025006/1025006/18.
Abstract
In the analysis of illposed inverse problems the impact of solution smoothness on accuracy and convergence rates plays an important role. For linear illposed operator equations in Hilbert spaces and with focus on the linear regularization schema we will establish relations between the different kinds of measuring solution smoothness in a pointwise or integral manner. In particular we discuss the interplay of distribution functions, profile functions that express the regularization error, index functions generating source conditions, and distance functions associated with benchmark source conditions. We show that typically the distance functions and the profile functions carry the same information as the distribution functions, and that this is not the case for general source conditions. The theoretical findings are accompanied with examples exhibiting applications and limitations of the approach. 
P. Mathé, U. Tautenhahn, Enhancing linear regularization to treat large noise, Journal of Inverse and IllPosed Problems, 19 (2011), pp. 859879.

P. Mathé, U. Tautenhahn, Regularization under general noise assumptions, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 27 (2011), pp. 035016/1035016/15.
Abstract
The authors explain how the major results which were obtained recently in Eggermont et al (2009 Inverse Problems 25 115018) can be derived from a more general perspective of recent regularization theory. By pursuing this further, the authors provide a general view on regularization under general noise assumptions, including weakly and strongly controlled noise. The prospect is not to generalize previous work in this direction, but rather to envision the intrinsic structure present in regularization under general noise assumptions. In particular, the authors find variants of the discrepancy and the Lepski$vrm i$ principle to choose the regularization parameter, albeit within different context and under different assumptions. 
D. Belomestny, Spectral estimation of the fractional order of a Lévy process, The Annals of Statistics, 38 (2010), pp. 317351.

B. Hoffmann, P. Mathé, H. VON Weizsäcker, Regularization in Hilbert space under unbounded operators and general source conditions, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 25 (2009), pp. 115013/1115013/15.

D. Belomestny, G.N. Milstein, V. Spokoiny, Regression methods in pricing American and Bermudan options using consumption processes, Quantitative Finance, 9 (2009), pp. 315327.
Abstract
Here we develop methods for efficient pricing multidimensional discretetime American and Bermudan options by using regression based algorithms together with a new approach towards constructing upper bounds for the price of the option. Applying sample space with payoffs at the optimal stopping times, we propose sequential estimates for continuation values, values of the consumption process, and stopping times on the sample paths. The approach admits constructing both low and upper bounds for the price by Monte Carlo simulations. The methods are illustrated by pricing Bermudan swaptions and snowballs in the Libor market model. 
P. Mathé, S.V. Pereverzev, The use of higher order finite difference schemes is not dangerous, Journal of Complexity, 25 (2009), pp. 310.

V. Spokoiny, C. Vial, Parameter tuning in pointwise adaptation using a propagation approach, The Annals of Statistics, 37 (2009), pp. 27832807.
Abstract
This paper discusses the problem of adaptive estimating a univariate object like the value of a regression function at a given point or a linear functional in a linear inverse problem. We consider an adaptive procedure originated from Lepski (1990) which selects in a datadriven way one estimate out of a given class of estimates ordered by their variability. A serious problem with using this and similar procedures is the choice of some tuning parameters like thresholds. Numerical results show that the theoretically recommended proposals appear to be too conservative and lead to a strong oversmoothing effects. A careful choice of the parameters of the procedure is extremely important for getting the reasonable quality of estimation. The main contribution of this paper is the new approach for choosing the parameters of the procedure by providing the prescribed behavior of the resulting estimate in the simple parametric situation. We establish a nonasymptotical “oracle” bound which shows that the estimation risk is, up to a logarithmic multiplier, equal to the risk of the “oracle” estimate which is optimally selected from the given family. A numerical study demonstrates the nice performance of the resulting procedure in a number of simulated examples. 
B. Hofmann, P. Mathé, M. Schieck, Modulus of continuity for conditionally stable illposed problems in Hilbert space, Journal of Inverse and IllPosed Problems, 16 (2008), pp. 567585.

P. Mathé, B. Hofmann, Direct and inverse results in variable Hilbert scales, Journal of Approximation Theory, 154 (2008), pp. 7789.

P. Mathé, B. Hofmann, How general are general source conditions?, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 24 (2008), pp. 015009/1015009/5.

P. Mathé, N. Schöne, Regularization by projection in variable Hilbert scales, Applicable Analysis. An International Journal, 2 (2008), pp. 201 219.

B. Hofmann, P. Mathé, S.V. Pereverzev, Regularization by projection: Approximation theoretic aspects and distance functions, Journal of Inverse and IllPosed Problems, 15 (2007), pp. 527545.

P. Mathé, U. Tautenhahn, Error bounds for regularization methods in Hilbert scales by using operator monotonicity, Far East Journal of Mathematical Sciences (FJMS), 24 (2007), pp. 121.
Abstract
For solving linear illposed problems with noisy data regularization methods are required. In the present paper regularized approximations are obtained by a general regularization scheme in Hilbert scales. We exploit operator monotonicity of certain functions for deriving order optimal error bounds that characterize the accuracy of the regularized approximations. These error bounds are obtained under general smoothness conditions 
P. Mathé, B. Hofmann, Analysis of profile functions for general linear regularization methods, SIAM Journal on Numerical Analysis, 45 (2007), pp. 11221141.
Abstract
The stable approximate solution of illposed linear operator equations in Hilbert spaces requires regularization. Tight bounds for the noisefree part of the regularization error are constitutive for bounding the overall error. Norm bounds of the noisefree part which decrease to zero along with the regularization parameter are called profile functions and are subject of our analysis. The interplay between properties of the regularization and certain smoothness properties of solution sets, which we shall describe in terms of sourcewise representations is crucial for the decay of associated profile functions. On the one hand, we show that a given decay rate is possible only if the underlying true solution has appropriate smoothness. On the other hand, if smoothness fits the regularization, then decay rates are easily obtained. If smoothness does not fit, then we will measure this in terms of some distance function. Tight bounds for these allow us to obtain profile functions. Finally we study the most realistic case when smoothness is measured with respect to some operator which is related to the one governing the original equation only through a link condition. In many parts the analysis is done on geometric basis, extending classical concepts of linear regularization theory in Hilbert spaces. We emphasize intrinsic features of linear illposed problems which are frequently hidden in the classical analysis of such problems. 
A. Goldenshluger, V. Spokoiny, Recovering convex edges of image from noisy tomographic data, IEEE Transactions on Information Theory, 52 (2006), pp. 13221334.

D. Belomestny, M. Reiss, Spectral calibration of exponential Lévy models, Finance and Stochastics, 10 (2006), pp. 449474.
Abstract
We investigate the problem of calibrating an exponential Lévy model based on market prices of vanilla options. We show that this inverse problem is in general severely illposed and we derive exact minimax rates of convergence. The estimation procedure we propose is based on the explicit inversion of the option price formula in the spectral domain and a cutoff scheme for high frequencies as regularisation. 
P. Mathé, U. Tautenhahn, Interpolation in variable Hilbert scales with application to inverse problems, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 22 (2006), pp. 22712297.
Abstract
For solving linear illposed problems with noisy data regularization methods are required. In the present paper regularized approximations in Hilbert scales are obtained by a general regularization scheme. The analysis of such schemes is based on new results for interpolation in Hilbert scales. Error bounds are obtained under general smoothness conditions. 
J. Polzehl, V. Spokoiny, Propagationseparation approach for local likelihood estimation, Probability Theory and Related Fields, 135 (2006), pp. 335362.
Abstract
The paper presents a unified approach to local likelihood estimation for a broad class of nonparametric models, including, e.g., regression, density, Poisson and binary response models. The method extends the adaptive weights smoothing (AWS) procedure introduced by the authors [Adaptive weights smoothing with applications to image sequentation. J. R. Stat. Soc., Ser. B 62, 335354 (2000)] in the context of image denoising. The main idea of the method is to describe a greatest possible local neighborhood of every design point in which the local parametric assumption is justified by the data. The method is especially powerful for model functions having large homogeneous regions and sharp discontinuities. The performance of the proposed procedure is illustrated by numerical examples for density estimation and classification. We also establish some remarkable theoretical nonasymptotic results on properties of the new algorithm. This includes the “propagation” property which particularly yields the root$n$ consistency of the resulting estimate in the homogeneous case. We also state an “oracle” result which implies rate optimality of the estimate under usual smoothness conditions and a “separation” result which explains the sensitivity of the method to structural changes. 
D. Belomestny, Reconstruction of the general distribution from the distribution of some statistics, Theory of Probability and its Applications, 49 (2005), pp. 115.
Abstract
We investigate the problem of characterizing the distribution of independent identically distributed random variables $X_1,ldots,X_m$ (general distribution) by the distribution of linear statistics and statistics of maximum with positive coefficients. Necessary and sufficient conditions are found under which such a characterization takes place. 
A. Goldenshluger, V. Spokoiny, On the shapefrommoments problem and recovering edges from noisy Radon data, Probability Theory and Related Fields, 128 (2004), pp. 123140.

D. Belomestny, Constraints on distributions imposed by properties of linear forms, ESAIM. Probability and Statistics, 7 (2003), pp. 313328.
Abstract
Let $(X_1,Y_1)...,(X_m,Y_m)$ be $m$ independent identically distributed bivariate vectors and $L_1=b_1X_1+...+b_mX_m$ ,$L_2=b_1Y_1+...+b_mY_m$ are two linear forms with positive coefficients. We study two problems: under what conditions does the equidistribution of $L_1$ and $L_2$ imply the same property for $X_1$ and $Y_1$, and under what conditions does the independence of $L_1$ and $L_2$ entail independence of $X_1$ and $Y_1$? Some analytical sufficient conditions are obtained and it is shown that in general they can not be weakened. 
P. Mathé, S.V. Pereverzev, Discretization strategy for linear illposed problems in variable Hilbert scales, Inverse Problems. An International Journal on the Theory and Practice of Inverse Problems, Inverse Methods and Computerized Inversion of Data, 19 (2003), pp. 12631277.
Abstract
The authors study the regularization of projection methods for solving linear illposed problems with compact and injective linear operators in Hilbert spaces. Smoothness of the unknown solution is given in terms of general source conditions, such that the framework of variable Hilbert scale s is suitable. The structure of the error is analyzed in terms of the noise level, the regularization parameter and as a function of other parameters, driving the discretization. As a result, a strategy is proposed, which automatically adapts to the unknown source condition, uniformly for certain classes, and provides the optimal order of accuracy. 
P. Mathé, S.V. Pereverzev, Direct estimation of linear functionals from indirect noisy observations, Journal of Complexity, 18 (2002), pp. 500516.
Abstract
The authors study the efficiency of the linear functional strategy, as introduced by Anderssen (1986), for inverse problems with observations blurred by Gaussian white noise with known intensity $delta$. The optimal accuracy is presented and it is shown, how this can be achieved by a linearfunctional strategy based on the noisy observations. This optimal linearfunctional strategy is obtained from Tikhonov regularization of some dual problem. Next, the situation is treated, when only a finite number of noisy observations, given beforehand is available. Under appropriate smoothness assumptions best possible accuracy still can be attained, if the number of observations corresponds to the noise intensity in a proper way. It is also shown, that, at least asymptotically this number of observations cannot be reduced. 
P. Mathé, Stable summation of orthogonal series with noisy coefficients, Journal of Approximation Theory, 117 (2002), pp. 6680.
Abstract
We study the recovery of continuous functions from Fourier coefficients with respect to certain given orthonormal systems, blurred by noise. For deterministic noise this is a classical illposed problem. Emphasis is laid on a priori smoothness assumptions on the solution, which allows to apply regularization to reach the best possible accuracy. Results are obtained for systems obeying norm growth conditions. In the white noise setting mild additional assumptions have to be made to have accurate bounds. We finish our study with the recovery of functions from noisy coefficients with respect to the Haar system. 
P. Mathé, S.V. Pereverzev, Optimal discretization of inverse problems in Hilbert scales. Regularization and selfregularization of projection methods, SIAM Journal on Numerical Analysis, 38 (2001), pp. 19992021.
Abstract
We study the efficiency of the approximate solution of illposed problems, based on discretized observations, which we assume to be given aforehand. We restrict ourselves to problems which can be formulated in Hilbert scales. Within this framework we shall quantify the degree of illposedness, provide general conditions on projection schemes to achieve the best possible order of accuracy. We pay particular attention on the problem of selfregularization vs. Tikhonov regularization. Moreover, we study the information complexity. Asymptotically, any method, which achieves the best possible order of accuracy must use at least such amount of noisy observations. We accomplish our study with two specific problems, Abel's integral equation and the recovery of continuous functions from noisy coefficients with respect to a given orthonormal system, both classical illposed problems.
Beiträge zu Sammelwerken

P. Dvurechensky, D. Dvinskikh, A. Gasnikov, C.A. Uribe, A. Nedić, Decentralize and randomize: Faster algorithm for Wasserstein barycenters, in: Advances in Neural Information Processing Systems 31, S. Bengio, H. Wallach, H. Larochelle, K. Grauman, N. CesaBianchi, R. Garnett, eds., Curran Associates, Inc., 2018, pp. 1076010770.

L. Bogulubsky, P. Dvurechensky, A. Gasnikov, G. Gusev, Y. Nesterov, A. Raigorodskii, A. Tikhonov, M. Zhukovskii, Learning supervised PageRank with gradientbased and gradientfree optimization methods, in: Advances in Neural Information Processing Systems 29, D.D. Lee, M. Sugiyama, U.V. Luxburg, I. Guyon, R. Garnett, eds., Curran Associates, Inc., 2016, pp. 49074915.

A. Chernov, P. Dvurechensky, A. Gasnikov, Fast primaldual gradient method for strongly convex minimization problems with linear constraints, in: Discrete Optimization and Operations Research  9th International Conference, DOOR 2016, Vladivostok, Russia, September 1923, 2016, Proceedings, Y. Kochetov, M. Khachay, V. Beresnev, E. Nurminski, P. Pardalos, eds., 9869 of Theoretical Computer Science and General Issues, Springer International Publishing Switzerland, Cham, 2016, pp. 391403.

M. Arias Chao, D.S. Lilley, P. Mathé, V. Schlosshauer, Calibration and uncertainty quantification of gas turbines performance models, in: ASME Turbo Expo 2015: Turbine Technical Conference and Exposition, Volume 7A: Structures and Dynamics, ASME and Alstom Technologie AG, 2015, pp. V07AT29A001V07AT29A012.

G. Blanchard, N. Krämer, Kernel partial least squares is universally consistent, in: Proceedings of the 13th International Conference on Artificial Intelligence and Statistics (AISTATS 2010), Y.W. Teh, M. Titterington, eds., 9 of JMLR Workshop and Conference Proceedings, Journal of Machine Learning Research, Cambridge, MA, USA, 2010, pp. 5764.
Preprints, Reports, Technical Reports

A. Ivanova, A. Gasnikov, P. Dvurechensky, D. Dvinskikh, A. Tyurin, E. Vorontsova, D. Pasechnyuk, Oracle complexity separation in convex optimization, Preprint no. 2711, WIAS, Berlin, 2020, DOI 10.20347/WIAS.PREPRINT.2711 .
Abstract, PDF (424 kByte)
Ubiquitous in machine learning regularized empirical risk minimization problems are often composed of several blocks which can be treated using different types of oracles, e.g., full gradient, stochastic gradient or coordinate derivative. Optimal oracle complexity is known and achievable separately for the full gradient case, the stochastic gradient case, etc. We propose a generic framework to combine optimal algorithms for different types of oracles in order to achieve separate optimal oracle complexity for each block, i.e. for each block the corresponding oracle is called the optimal number of times for a given accuracy. As a particular example, we demonstrate that for a combination of a full gradient oracle and either a stochastic gradient oracle or a coordinate descent oracle our approach leads to the optimal number of oracle calls separately for the full gradient part and the stochastic/coordinate descent part.
Vorträge, Poster

M. Hintermüller, A function space framework for structural total variation regularization with applications in inverse problems, 71st Workshop: Advances in Nonsmooth Analysis and Optimization (NAO2019), June 25  30, 2019, International School of Mathematics ``Guido Stampacchia'', Erice, Italy, June 26, 2019.

M. Hintermüller, A function space framework for structural total variation regularization with applications in inverse problems, Thematic Programme ``Modern Maximal Monotone Operator Theory: From Nonsmooth Optimization to Differential Inclusions'', Workshop ``Nonsmooth and Variational Analysis'', January 28  February 1, 2019, Erwin Schrödinger International Institute for Mathematics and Physics, Vienna, Austria, February 1, 2019.

M. Hintermüller, Applications in image processing, Workshop on Efficient Operator Splitting Techniques for Complex System and Large Scale Data Analysis, January 15  18, 2019, Sanya, China, January 14, 2019.

M. Hintermüller, Structural total variation regularization with applications in inverse problems, 90th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM), DFG Priority Programme 1962 ``Non Smooth and ComplementarityBased Distributed Parameter Systems: Simulation and Hierarchical Optimization'', February 18  22, 2019, Technische Universität Wien, Austria, February 19, 2019.

M. Marschall, Adaptive lowrank approximation in Bayesian inverse problems, 3rd International Conference on Uncertainty Quantification in Computational Sciences and Engineering (UNCECOMP 2019), Minisymposium 6IV ``Uncertainty Computations with Reduced Order Models and LowRank Representations'', June 24  26, 2019, Crete, Greece, June 25, 2019.

M. Marschall, Complexity reduction in Bayesian inverse problems by lowrank tensor representation, Robert Bosch GmbH, Corporate Research  Advanced Engineering Computer Vision Systems (CR/AEC4), Hildesheim, April 16, 2019.

M. Marschall, Lowrank surrogates in Bayesian inverse problems, 19th FrenchGermanSwiss Conference on Optimization (FGS'2019), Minisymposium 1: ``Recent Trends in Nonlinear Optimization 1'', September 17  20, 2019, Nice, France, September 17, 2019.

M. Marschall, Random domains in PDE problems with lowrank surrogates. Forward and backward, PhysikalischTechnische Bundesanstalt, Berlin, April 10, 2019.

P. Mathé, Relating direct and inverse Bayesian problems via the modulus of continue, Stochastic Computation and Complexity, April 15  16, 2019, Institut Henri Poincaré, Paris, France, April 16, 2019.

P. Mathé, Relating direct and inverse problems via the modulus of continuity, The Chemnitz Symposium on Inverse Problems 2019, September 30  October 2, 2019, TU Chemnitz, Fakultät für Mathematik, Frankfurt a. M., October 1, 2019.

P. Mathé, The role of the modulus of continuity in inverse problems, Forschungsseminar Inverse Probleme, Technische Universität Chemnitz, Fachbereich Mathematik, August 13, 2019.

M. Eigel, Adaptive Galerkin FEM for stochastic forward and inverse problems, Optimisation and Numerical Analysis Seminars, University of Birmingham, School of Mathematics, UK, February 15, 2018.

M. Eigel, Adaptive tensor methods for forward and inverse problems, SIAM Conference on Uncertainty Quantification (UQ18), Minisymposium 122 ``LowRank Approximations for the Forward and the Inverse Problems III'', April 16  19, 2018, Garden Grove, USA, April 19, 2018.

M. Hintermüller, Automated regularization parameter choice rule in image processing, Workshop ``New Directions in Stochastic Optimisation'', August 19  25, 2018, Mathematisches Forschungsinstitut Oberwolfach, August 23, 2018.

M. Hintermüller, Bilevel optimisation in automated regularisation parameter selection in image processing, WIASPGMO Workshop on Nonsmooth and Stochastic Optimization, June 26, 2018, HumboldtUniversität zu Berlin, June 26, 2018.

M. Hintermüller, Bilevel optimization and some "parameter learning" applications in image processing, SIAM Conference on Imaging Science, Minisymposium MS5 ``Learning and Adaptive Approaches in Image Processing'', June 5  8, 2018, Bologna, Italy, June 5, 2018.

M. Marschall, Bayesian inversion with adaptive lowrank approximation, Analysis, Control and Inverse Problems for PDEs  Workshop of the FrenchGermanItalian LIA (Laboratoire International Associe) COPDESC on Applied Analysis, November 26  30, 2018, University of Naples Federico II and Accademia Pontaniana, Italy, November 29, 2018.

P. Mathé, Bayesian inverse problems with noncommuting operators, Chemnitz September of Applied Mathematics 2018, Chemnitz Symposium on Inverse Problems, September 27  28, 2018, Technische Universität Chemnitz, Fakultät für Mathematik, September 28, 2018.

P. Mathé, Complexity of linear illposed problems in Hilbert space, Stochastisches Kolloquium, GeorgAugustUniversität Göttingen, Institut für Mathematische Stochastik, February 7, 2018.

M. Eigel, Adaptive stochastic FE for explicit Bayesian inversion with hierarchical tensor representations, Institut National de Recherche en Informatique et en Automatique (INRIA), SERENA (Simulation for the Environment: Reliable and Efficient Numerical Algorithms) research team, Paris, France, June 1, 2017.

R. Gruhlke, Multiscale failure analysis with polymorphic uncertainties for optimal design of rotor blades, Frontiers of Uncertainty Quantification in Engineering (FrontUQ 2017), September 6  8, 2017, München, September 6, 2017.

M. Hintermüller, Bilevel optimization and applications in imaging, Workshop ``Emerging Developments in Interfaces and Free Boundaries'', January 22  28, 2017, Mathematisches Forschungsinstitut Oberwolfach.

M. Hintermüller, Bilevel optimization and applications in imaging, Mathematisches Kolloquium, Universität Wien, Austria, January 18, 2017.

M. Hintermüller, Bilevel optimization and some ``parameter learning'' applications in image processing, LMS Workshop ``Variational Methods Meet Machine Learning'', September 18, 2017, University of Cambridge, Centre for Mathematical Sciences, UK, September 18, 2017.

A. Koziuk, Bootstrap for the regression problem with instrumental variables, Haindorf Seminar 2017, January 24  28, 2017, HumboldtUniversität zu Berlin, Wirtschaftswissenschaftliche Fakultät, Hejnice, Czech Republic, January 26, 2017.

M. Marschall, Bayesian inversion using hierarchical tensors, 88th Annual Meeting of the International Association of Applied Mathematics and Mechanics (GAMM 2017), Section S15 ``Uncertainty Quantification'', March 6  10, 2017, Bauhaus Universität Weimar/Technische Universität Ilmenau, Weimar, March 8, 2017.

M. Marschall, Samplingfree Bayesian inversion with adaptive hierarchical tensor representation, Frontiers of Uncertainty Quantification in Engineering (FrontUQ 2017), September 6  8, 2017, München, September 7, 2017.

M. Marschall, Samplingfree Bayesian inversion with adaptive hierarchical tensor representation, International Conference on Scientific Computation and Differential Equations (SciCADE2017), MS21 ``Tensor Approximations of MultiDimensional PDEs'', September 11  15, 2017, University of Bath, UK, September 14, 2017.

P. Mathé, Bayesian inverse problems with noncommuting operators, Statistical Foundations of Uncertainty Quantification for Inverse Problems Workshop, June 19  22, 2017, University of Cambridge, Center for Mathematical Sciences, UK, June 21, 2017.

P. Mathé, Complexity of supervised learning, ibcparis2017 : Information Based Complexity, HighDimensional Problems, March 14  15, 2017, Institut Henri Poincaré, Paris, France, March 15, 2017.

P. Mathé, Numerical integration (mini course), November 20  December 4, 2017, Fudan University, School of Mathematical Sciences, China.

N. Buzun, Multiplier bootstrap for change point detection, Mathematical Statistics and Inverse Problems, February 8  12, 2016, Faculté des Sciences de Luminy, France, February 11, 2016.

N. Buzun, Multiplier bootstrap for change point detection, Spring School ``Structural Inference 2016", March 13  18, 2016, DFG Forschergruppe FOR 1735, Lübeck, Germany, March 14, 2016.

M. Eigel, Bayesian inversion using hierarchical tensor approximations, SIAM Conference on Uncertainty Quantification, Minisymposium 67 ``Bayesian Inversion and Lowrank Approximation (Part II)'', April 5  8, 2016, Lausanne, Switzerland, April 6, 2016.

T. Wu, Bilevel optimization and applications in imaging sciences, August 24  25, 2016, Shanghai Jiao Tong University, Institute of Natural Sciences, China.

P. Dvurechensky, Gradient and gradientfree methods for pagerank algorithm learning, Workshop on Modern Statistics and Optimization, February 23  24, 2016, Russian Academy of Sciences, Institute for Information Transmission Problems, Moscow, Russian Federation, February 24, 2016.

P. Dvurechensky, Random gradientfree methods for webpage ranking model learning, 30th annual conference of the Belgian Operational Research Society, January 27  29, 2016, LouvainlaNeuve, Belgium, January 28, 2016.

M. Hintermüller, K. Papafitsoros, C. Rautenberg, A fine scale analysis of spatially adapted total variation regularisation, Imaging, Vision and Learning based on Optimization and PDEs, Bergen, Norway, August 29  September 1, 2016.

M. Hintermüller, Bilevel optimization and applications in imaging, Imaging, Vision and Learning based on Optimization and PDEs, August 29  September 1, 2016, Bergen, Norway, August 30, 2016.

M. Hintermüller, Bilevel optimization for a generalized totalvariation model, SIAM Conference on Imaging Science, Minisymposium ``NonConvex Regularization Methods in Image Restoration'', May 23  26, 2016, Albuquerque, USA, May 26, 2016.

M. Hintermüller, Optimal selection of the regularisation function in a localised TV model, SIAM Conference on Imaging Science, Minisymposium ``Analysis and Parameterisation of Derivative Based Regularisation'', May 23  26, 2016, Albuquerque, USA, May 24, 2016.

P. Mathé, Complexity of linear illposed problems in Hilbert space, IBC on the 70th anniversary of Henryk Wozniakowski, August 29  September 2, 2016, Banach Center, Bedlewo, Poland, August 31, 2016.

P. Mathé, Complexity of linear illposed problems in Hilbert space, Chemnitz Symposium on Inverse Problems, September 22  23, 2016, Technische Universität Chemnitz, Fakultät für Mathematik, September 22, 2016.

P. Mathé, Discrepancy based model selection in statistical inverse problems, Mathematical Statistics and Inverse Problems, February 8  12, 2016, Centre International de Rencontres Mathématiques (CIRM), Luminy, France, February 11, 2016.

V. Spokoiny, Gradient and gradientfree methods for pagerank algorithm learning, Workshop on Modern Statistics and Optimization, February 23  24, 2016, Russian Academy of Sciences, Institute for Information Transmission Problems, Moscow, Russian Federation.

P. Dvurechensky, Semisupervised pagerank model learning with gradientfree optimization methods, Traditional Youth School ``Control, Information and Optimization'', June 14  20, 2015, Moscow, Russian Federation, June 17, 2015.

P. Mathé, Minimax signal detection in statistical inverse problems, Algorithms and Complexity for Continuous Problems, September 21  25, 2015, Schloss Dagstuhl, September 25, 2015.

P. Mathé, A random surfer in the internet, GermanEstonian Academic Week ``Academica'', October 13  15, 2014, University of Tartu, Faculty of Mathematics and Computer Science, Estonia, October 14, 2014.

P. Mathé, Bayesian analysis of statistical inverse problems, Colloquium of the Faculty of Mathematics and Computer Science, University of Tartu, Estonia, October 13, 2014.

P. Mathé, Bayesian regularization of statistical inverse problems, Rencontres de Statistique Mathématique: Nouvelles Procédures pour Nouvelles Données, December 15  19, 2014, Centre International de Rencontres Mathématiques (CIRM), Marseille, France, December 17, 2014.

P. Mathé, Convergence of Bayesian schemes in inverse problems, Kolloquium über Angewandte Mathematik, GeorgAugustUniversität Göttingen, Fakultät für Mathematik und Informatik, June 10, 2014.

P. Mathé, Merging regularization theory into Bayesian inverse problems, Chemnitz Symposium on Inverse Problems, September 15  19, 2014, Universität Chemnitz, Fachbereich Mathematik, September 18, 2014.

P. Mathé, Oracletype posterior contraction rates in Bayesian inverse problems, International Workshop ``Advances in Optimization and Statistics'', May 15  16, 2014, Russian Academy of Sciences, Institute of Information Transmission Problems (Kharkevich Institute), Moscow, May 16, 2014.

P. Mathé, Parameter estimation by thresholding, Mathematical Modelling and Simulation Workshop, WIAS Berlin, April 15, 2014.

P. Mathé, Smoothness beyond differentiability, Seminar for Doctoral Candidates, University of Tartu, Faculty of Mathematics and Computer Science, Estonia, October 15, 2014.

A. Andresen, Finite sample analysis of maximum likelihood estimators and convergence of the alternating procedure, 29th European Meeting of Statisticians (EMS), July 20  25, 2013, Eötvös Loránd University, Budapest, Hungary, July 20, 2013.

H. Mai, Estimating a subordinators density, DynStoch 2013, April 17  19, 2013, University of Copenhagen, Department of Mathematical Sciences, Denmark, April 17, 2013.

P. Mathé, Bayes analysis for model calibration, Alstom Workshop, WIAS, May 7, 2013.

P. Mathé, Signal detection in inverse problems, PreMoLab Workshop on: Advances in predictive modeling and optimization, May 16  17, 2013, WIASBerlin, May 16, 2013.

P. Mathé, Signal detection in inverse problems, Mathematical Modelling and Analysis ( MMA2013 ) and Approximation Methods and Orthogonal Expansions ( AMOE2013 ), May 27  31, 2013, University of Tartu, Institute of Mathematics, Estonia, May 29, 2013.

P. Mathé, Statistical Inverse Problems, Applied Math Seminar, University of Warwick, Mathematics Institute, Coventry, UK, October 18, 2013.

S. Becker, Image processing via orientation scores, Workshop ``Computational Inverse Problems'', October 23  26, 2012, Mathematisches Forschungsinstitut Oberwolfach, October 25, 2012.

A. Andresen, Non asymptotic Wilks phenomenon in semiparametric estimation, PreMoLab: MoscowBerlin Stochastic and Predictive Modeling, May 31  June 1, 2012, Russian Academy of Sciences, Institute for Information Transmission Problems (Kharkevich Institute), Moscow, May 31, 2012.

A. Andresen, Nonasymptotic Wilks phenomenon in semiparametric estimation, 2. Structural Inference Day, WIAS, April 23, 2012.

A. Andresen, Nonasymptotic Wilks phenomenon in semiparametric estimation, Haindorf Seminar 2012 (Klausurtagung des SFB 649), February 9  12, 2012, HumboldtUniversität zu Berlin, Wirtschaftswissenschaftliche Fakultät, Hejnice, Czech Republic, February 10, 2012.

P. Mathé, An oracletype bound for a statistical RGrule, Nonparametric and Highdimensional Statistics, December 17  21, 2012, Centre International de Rencontres Mathématiques (CIRM), Marseille, France, December 20, 2012.

P. Mathé, Diskrepanzbasierte Parameterwahl in statistischen inversen Problemen, Technische Universität Chemnitz, Fakultät für Mathematik, September 19, 2012.

P. Mathé, Regularization of statistical inverse problems in Hilbert space, Journées Statistiques du Sud 2012, June 20  22, 2012, Université Toulouse, Institut National des Sciences Appliquées, France, June 20, 2012.

P. Mathé, Statistische Datenanalyse unter BOP, Workshop on Simulation in Industrial Process Engineering, WIAS, September 6, 2012.

P. Mathé, Using the discrepancy principle in statistical inverse problems, Regularisation symposium, Australian National University, Mathematical Sciences Institute, Canberra, November 22, 2012.

V. Spokoiny, Basics of modern parametric statistics, February 13  28, 2012, Independent University of Moscow, Center for Continuous Mathematical Education, Russian Federation.

V. Spokoiny, Bernsteinvon Mises theorem for quasi posteriors, International Workshop on Recent Advances in Time Series Analysis (RATS 2012), June 8  12, 2012, University of Cyprus, Department of Mathematics and Statistics, Protaras, June 9, 2012.

V. Spokoiny, Bernsteinvon Mises theorem for quasi posteriors, Workshop ``Frontiers in Nonparametric Statistics'', March 11  17, 2012, Mathematisches Forschungsinstitut Oberwolfach, March 12, 2012.

V. Spokoiny, Bernsteinvon Mises theorem for quasi posteriors, Workshop II on Financial Time Series Analysis: Highdimensionality, Nonstationarity and the Financial Crisis, June 19  22, 2012, National University of Singapore, Institute for Mathematical Sciences, June 21, 2012.

V. Spokoiny, Bernsteinvon Mises theorem for quasi posteriors, Workshop on Recent Developments in Statistical Multiscale Methods, July 16  18, 2012, GeorgAugustUniversität Göttingen, Institut für Mathematische Stochastik, July 17, 2012.

V. Spokoiny, Bernsteinvon Mises theorem for quasi posteriors, PreMoLab Seminar, Russian Academy of Sciences, Institute for Information Transmission Problems (Kharkevich Institute), Moscow, March 15, 2012.

V. Spokoiny, Parametric estimation: Modern view, PreMoDay I, February 24, 2012, Russian Academy of Sciences, Institute for Information Transmission Problems (Kharkevich Institute), Moscow, February 24, 2012.

V. Spokoiny, Some methods of modern statistics, Information Technology and Systems 2, August 19  25, 2012, Russian Academy of Sciences, Institute for Information Transmission Problems (Kharkevich Institute), Petrozavodsk, August 20, 2012.

P. Mathé, Conjugate gradient iteration with noisy data, Foundations of Computational Mathematics (FoCM'11), July 4  14, 2011, Budapest, Hungary, July 6, 2011.

V. Spokoiny, Alternating and semiparametric efficiency, École Nationale de la Statistique et de l'Analyse de l'Information (ENSAI), Rennes, France, September 16, 2011.

V. Spokoiny, Modern parametric theory, September 13  16, 2011, École Nationale de la Statistique et de l'Analyse de l'Information (ENSAI), Rennes, France.

P. Mathé, Conjugate gradient iteration under white noise, International Conference on Scientific Computing (SC2011), October 10  14, 2011, Università di Cagliari, Dipartimento di Matematica e Informatica, Cagliari, Italy, October 14, 2011.

V. Spokoiny, Semiparametric estimation alternating and efficiency, École Nationale de la Statistique et de l'Administration Économique (ENSAE), Paris, France, December 12, 2011.

M. Becker, Selfintersection local times: Exponential moments in subcritical dimensions, Excess SelfIntersections and Related Topics, December 6  10, 2010, Centre international de rencontres mathématiques (CIRM), Luminy, France, December 6, 2010.

S. Becker, Regularization of statistical inverse problems and the Bakushinskii veto, Rencontres de Statistiques Mathématiques 10, December 13  17, 2010, Centre International de Rencontres Mathématiques (CIRM), Luminy, France, December 17, 2010.

N. Krämer, Conjugate gradient regularization  A statistical framework for partial least squares regression, 4th Workshop on Partial Least Squares and Related Methods for Cuttingedge Research in Experimental Sciences, May 10  11, 2010, École Supérieure d'Électricité, Department of Information Systems and Decision Sciences, GifsurYvette, France, May 11, 2010.

D. Belomestny, Estimating the distribution of jumps in regular affine models: Uniform rates of convergence, Leipziger StochastikTage, March 1  5, 2010, Universität Leipzig, Fakultät für Mathematik und Informatik, March 2, 2010.

D. Belomestny, Statistical inference for multidimensional timechanged Levy processes based on lowfrequency data, 28th European Meeting of Statisticians, August 19  23, 2010, University of Piraeus, Department of Statistics and Insurance Science, Greece, August 23, 2010.

G. Blanchard, N. Krämer, Kernel partial least squares is universally consistent, AI & Statistics 2010, Sardinia, Italy, May 13  15, 2010.

G. Blanchard, N. Krämer, Optimal rates for conjugate gradient regularization, AI & Statistics 2010, Sardinia, Italy, May 13  15, 2010.

P. Mathé, Analysis of inverse problems under general smoothness assumptions, 5th International Conference on Inverse Problems: Modeling and Simulation, May 24  29, 2010, Izmir University, Department of Mathematics and Computer Sciences, Antalya, Turkey, May 25, 2010.

P. Mathé, Regularization under general noise assumptions, Chemnitz Symposium on Inverse Problems 2010, September 23  24, 2010, Chemnitz University of Technology, Department of Mathematics, September 23, 2010.

P. Mathé, Warum und wie rechnen wir mit allgemeinen Quelldarstellungen?, Seminar Inverse Probleme (Fr. C. Böckmann), Universität Potsdam, Institut für Mathematik, January 12, 2010.

D. Belomestny, Estimation of the jump activity of a Lévy process from low frequency data, Haindorf Seminar 2009, February 12  15, 2009, HumboldtUniversität zu Berlin, CASE  Center for Applied Statistics and Economics, Hejnice, Czech Republic, February 12, 2009.

D. Belomestny, Spectral estimation of the fractional order of a Lévy process, Workshop ``Statistical Inference for Lévy Processes with Applications to Finance'', July 15  17, 2009, EURANDOM, Eindhoven, Netherlands, July 16, 2009.

G. Blanchard, Convergence du gradient conjugué fonctionnel pour l'apprentissage statistique, École Normale Supérieure, Paris, France, March 16, 2009.

P. Mathé, Approximation theoretic aspects in variable Hilbert scales, 2nd Dolomites Workshop on Constructive Approximation and Applications, September 3  10, 2009, Università degli Studi di Verona, Dipartimento di Informatica, Italy, September 5, 2009.

P. Mathé, Discretization under general smoothness assumptions, Monday Lecture Series, Johann Radon Institute for Computational and Applied Mathematics (RICAM), Linz, Austria, June 29, 2009.

P. Mathé, Inverse problems in different settings, Seminar ``Algorithms and Complexity for Continuous Problems'', September 20  25, 2009, Schloss Dagstuhl, September 25, 2009.

P. Mathé, On some minimizationbased heuristic parameter choice in inverse problems, ChemnitzRICAM Symposium on Inverse Problems, Johann Radon Institute for Computational and Applied Mathematics (RICAM), Linz, Austria, July 14, 2009.

P. Mathé, Typical behavior in heuristic parameter choice, KickOff Meeting of Mini Special Semester on Inverse Problems, May 18  22, 2009, Johann Radon Institute for Computational and Applied Mathematics (RICAM), Linz, Austria, May 18, 2009.

V. Spokoiny, Saddle point model selection for inverse problems, Workshop ``Challenges in Statistical Theory: Complex Data Structures and Algorithmic Optimization'', August 23  29, 2009, Mathematisches Forschungsinstitut Oberwolfach, August 26, 2009.

P. Mathé, Average case analysis of inverse problems, Symposium on Inverse Problems 2008, September 25  26, 2008, TU Chemnitz, Fachbereich Mathematik, September 25, 2008.
Preprints im Fremdverlag

A. Ivanova, A. Gasnikov, P. Dvurechensky, D. Dvinskikh, A. Tyurin, E. Vorontsova, D. Pasechnyuk, Oracle complexity separation in convex optimization, Preprint no. arXiv:2002.02706, Cornell University, .
Abstract
Ubiquitous in machine learning regularized empirical risk minimization problems are often composed of several blocks which can be treated using different types of oracles, e.g., full gradient, stochastic gradient or coordinate derivative. Optimal oracle complexity is known and achievable separately for the full gradient case, the stochastic gradient case, etc. We propose a generic framework to combine optimal algorithms for different types of oracles in order to achieve separate optimal oracle complexity for each block, i.e. for each block the corresponding oracle is called the optimal number of times for a given accuracy. As a particular example, we demonstrate that for a combination of a full gradient oracle and either a stochastic gradient oracle or a coordinate descent oracle our approach leads to the optimal number of oracle calls separately for the full gradient part and the stochastic/coordinate descent part. 
P. Dvurechensky, S. Shtern, M. Staudigl, P. Ostroukhov, K. Safin, Selfconcordant analysis of FrankWolfe algorithms, Preprint no. arXiv:2002.04320, Cornell University, .
Abstract
Projectionfree optimization via different variants of the FrankWolfe (FW) method has become one of the cornerstones in optimization for machine learning since in many cases the linear minimization oracle is much cheaper to implement than projections and some sparsity needs to be preserved. In a number of applications, e.g. Poisson inverse problems or quantum state tomography, the loss is given by a selfconcordant (SC) function having unbounded curvature, implying absence of theoretical guarantees for the existing FW methods. We use the theory of SC functions to provide a new adaptive step size for FW methods and prove global convergence rate O(1/k), k being the iteration counter. If the problem can be represented by a local linear minimization oracle, we are the first to propose a FW method with linear convergence rate without assuming neither strong convexity nor a Lipschitz continuous gradient. 
A. Rastogi, G. Blanchard, P. Mathé, Convergence analysis of Tikhonov regularization for nonlinear statistical inverse learning problems, Preprint no. arXiv:1902.05404, Cornell University Library, arXiv.org, 2019.
Abstract
We study a nonlinear statistical inverse learning problem, where we observe the noisy image of a quantity through a nonlinear operator at some random design points. We consider the widely used Tikhonov regularization (or method of regularization, MOR) approach to reconstruct the estimator of the quantity for the nonlinear illposed inverse problem. The estimator is defined as the minimizer of a Tikhonov functional, which is the sum of a data misfit term and a quadratic penalty term. We develop a theoretical analysis for the minimizer of the Tikhonov regularization scheme using the ansatz of reproducing kernel Hilbert spaces. We discuss optimal rates of convergence for the proposed scheme, uniformly over classes of admissible solutions, defined through appropriate source conditions. 
F. Stonyakin, A. Gasnikov, A. Tyurin, D. Pasechnyuk, A. Agafonov, P. Dvurechensky, D. Dvinskikh, A. Kroshnin, V. Piskunova, Inexact model: A framework for optimization and variational inequalities, Preprint no. arXiv:1902.00990, Cornell University Library, arXiv.org, 2019.

C.A. Uribe, D. Dvinskikh, P. Dvurechensky, A. Gasnikov, A. Nedić, Distributed computation of Wasserstein barycenters over networks, Preprint no. arXiv:1803.02933, Cornell University Library, arXiv.org, 2018.

P. Dvurechensky, D. Dvinskikh, A. Gasnikov, C.A. Uribe, A. Nedić, Decentralize and randomize: Faster algorithm for Wasserstein barycenters, Preprint no. arXiv:1806.03915, Cornell University Library, arXiv.org, 2018.
Abstract
We study the problem of decentralized distributed computation of a discrete approximation for regularized Wasserstein barycenter of a finite set of continuous probability measures distributedly stored over a network. Particularly, we assume that there is a network of agents/machines/computers where each agent holds a private continuous probability measure, and seeks to compute the barycenter of all the measures in the network by getting samples from its local measure and exchanging information with its neighbors. Motivated by this problem, we develop and theoretically analyze a novel accelerated primaldual stochastic gradient method for general stochastic convex optimization problems with linear equality constraints. Then, we apply this method to the decentralized distributed optimization setting to propose a new algorithm for the distributed semidiscrete regularized Wasserstein barycenter problem. The proposed algorithm can be executed over arbitrary networks that are undirected, connected and static, using the local information only. Moreover, we show explicit nonasymptotic complexity in terms of the problem parameters. Finally, we show the effectiveness of our method on the distributed computation of the regularized Wasserstein barycenter of univariate Gaussian and von Mises distributions, as well as on some applications to image aggregation. 
P. Mathé, Bayesian inverse problems with noncommuting operators, Preprint no. arXiv:1801.09540, Cornell University Library, arXiv.org, 2018.
Abstract
The Bayesian approach to illposed operator equations in Hilbert space recently gained attraction. In this context, and when the prior distribution is Gaussian, then two operators play a significant role, the one which governs the operator equation, and the one which describes the prior covariance. Typically it is assumed that these operators commute. Here we extend this analysis to noncommuting operators, replacing the commutativity assumption by a link condition. We discuss its relation to the commuting case, and we indicate that this allows to use interpolation type results to obtain tight bounds for the contraction of the posterior Gaussian distribution towards the data generating element. 
M. Hintermüller, N. Strogies, On the identification of the friction coefficient in a semilinear system for gas transport through a network, Preprint, DFG SFB Transregio 154 ``Mathematical Modelling, Simulation and Optimization using the Example of Gas Networks'', 2017.

L. Bogolubsky, P. Dvurechensky, A. Gasnikov, G. Gusev, Y. Nesterov, A. Raigorodskii, A. Tikhonov, M. Zhukovskii, Learning supervised PageRank with gradientbased and gradientfree optimization methods, Preprint no. arXiv:1603.00717, Cornell University Library, arXiv.org, 2016.

A. Chernov, P. Dvurechensky, A. Gasnikov, Fast primaldual gradient method for strongly convex minimization problems with linear constraints, Preprint no. arXiv:1605.02970, Cornell University Library, arXiv.org, 2016.

S.W. Anzengruber, B. Hofmann, P. Mathé, Regularization properties of the discrepancy principle for Tikhonov regularization in Banach spaces, Preprint no. 12, Technische Universität Chemnitz, Fakultät für Mathematik, 2012.

B. Hofmann, P. Mathé, Parameter choice in Banach space regularization under variational inequalities, Preprint no. 05, Technische Universität Chemnitz, Fakultät für Mathematik, 2012.

V. Spokoiny, Parametric estimation. Finite sample theory, Preprint no. arXiv:1111.3029, Cornell University Library, arXiv.org, 2012.

V. Spokoiny, Roughness penalty, Wilks phenomenon, and Bernsteinvon Mises theorem, Preprint no. arXiv:1205.0498, Cornell University Library, arXiv.org, 2012.

R.I. Boţ, B. Hofmann, P. Mathé, Regularizability of illposed problems and the modulus of continuity, Preprint no. 17, Technische Universität Chemnitz, Fakultät für Mathematik, 2011.

B. Hofmann, P. Mathé, Some note on the modulus of continuity for illposed problems in Hilbert space, Preprint no. 07, Technische Universität Chemnitz, Fakultät für Mathematik, 2011.

G. Blanchard, P. Mathé, Discrepancy principle for statistical inverse problems with application to conjugate gradient iteration, Preprint no. 07, Universität Potsdam, Institut für Mathematik, 2011.