Darija Marković
Associate Professor Department of Mathematics Josip Juraj Strossmayer University of Osijek Trg Ljudevita Gaja 6 Osijek, HR31000, Croatia¸

Research Interests
 Applied Mathematics
 Numerical Mathematics
 Parameter Estimation
Degrees
 PhD in Numerical Mathematics, Faculty of Natural Science, Department of Mathematics, University of Zagreb, 2009.
 MSc in Mathematics, Faculty of Natural Science, Department of Mathematics, University of Zagreb, 2005.
 BSc in Mathematics and Computer Science, Department of Mathematics, University of Osijek, Croatia, 2000.
Publications
 R. Čorić, D. Matijević, D. Marković, PollenNet  a deep learning approach to predicting airborne pollen concentrations, Croatian Operational Research Review 14/1 (2023), 113The accurate shortterm forecasting of daily airborne pollen concentrations is of great importance in public health. Various machine learning and statistical techniques have been employed to predict these concentrations. In this paper, an RNNbased method called PollenNet is introduced, which is capable of predicting the average daily pollen concentrations for three types of pollen: ragweed (Ambrosia), birch (Betula), and grass (Poaceae). Moreover, two strategies incorporating measurement errors during the training phase are introduced, making the method more robust. The data for experiments were obtained from the RealForAll project, where pollen concentrations were gathered using a Hirsttype 7day volumetric spore trap. Additionally, five types of meteorological data were utilized as input variables. The results of our study demonstrate that the proposed method outperforms standard models typically used for predicting pollen concentrations, specifically the pollen calendar method, pollen predictions based on patterns, and the naive approach.
 D. Jukić, D. Marković, Nonlinear least squares estimation of the shifted Gompertz distribution, European Journal of Pure and Applied Mathematics 10/2 (2017), 157166The focus of this paper is the existence of the best nonlinear least squares estimate for the shifted Gompertz distribution. As a main result, two theorems on the existence of the least squares estimate are obtained, as well as their generalization in the lp norm (1 ≤ p < ∞ ).
 D. Marković, Preponderantly increasing/decreasing data in regression analysis, Croatian Operational Research Review 7/2 (2016), 269276For the given data (wi,xi,yi), i=1, ..., n, and the given model function f(x;θ), where θ is a vector of unknown parameters, the goal of regression analysis is to obtain estimator θ∗ of the unknown parameters θ such that the vector of residuals is minimized in some sense. The common approach to this problem of minimization is the leastsquares method, that is minimizing the L2 norm of the vector of residuals. For nonlinear model functions, what is necessary is finding at least the sufficient conditions on the data that will guarantee the existence of the best leastsquares estimator. In this paper we will describe and examine in detail the property of preponderant increase/decrease of the data, which ensures the existence of the best estimator for certain important nonlinear model functions.
 D. Marković, L. Borozan, On Parameter Estimation by Nonlinear Least Squares in Some Special TwoParameter Exponential Type Models, Applied Mathematics & Information Sciences 9/6 (2015), 29252931Twoparameter growth models of exponential type f(t;a,b) = g(t)exp(a+bh(t)), where a and b are unknown parameters and g and h are some known functions, are frequently employed in many different areas such as biology, finance, statistic, medicine, ect. The unknown parameters must be estimated from the data (w_i, t_i, y_i), i = 1,...,n, where t_i denote the values of the independent variable, y_i are respective estimates of regression function f and w_i > 0 are some data weights. A very popular and widely used method for parameter estimation is the method of least squares. In practice, to avoid using nonlinear regression, this kind of problems are commonly transformed to linear, which is not statistically justified. In this paper we show that for strictly positive g and strictly monotone h original nonlinear problem has a solution. Generalization in the lp norm (1 ≤ p < ∞) and some illustrative examples are also given.
 D. Marković, D. Jukić, Total least squares fitting the threeparameter inverse Weibull density, European Journal of Pure and Applied Mathematics 7/3 (2014), 230245The focus of this paper is on a nonlinear weighted total least squares fitting problem for the threeparameter inverse Weibull density which is frequently employed as a model in reliability and lifetime studies. As a main result, a theorem on the existence of the total least squares estimator is obtained, as well as its generalization in the l_q norm (1≤q<∞).
 D. Marković, D. Jukić, On parameter estimation in the bass model by nonlinear least squares fitting the adoption curve, International Journal of Applied Mathematics and Computer Science 23/1 (2013), 145155The Bass model is one of the most wellknown and widely used firstpurchase diffusion models in marketing research. Estimation of its parameters has been approached in the literature by various techniques. In this paper, we consider the parameter estimation approach for the Bass model based on nonlinear weighted least squares fitting of its derivative known as the adoption curve. We show that it is possible that the least squares estimate does not exist. As a main result, two theorems on the existence of the least squares estimate are obtained, as well as their generalization in the ls norm (1 ≤ s < ∞). One of them gives necessary and sufficient conditions which guarantee the existence of the least squares estimate. Several illustrative numerical examples are given to support the theoretical work.
 M. Marušić, D. Marković, D. Jukić, Least squares fitting the threeparameter inverse Weibull density, Mathematical Communications 15/2 (2010), 539553The inverse Weibull model was developed by Erto [10]. In practice, the unknown parameters of the ppropriate inverse Weibull density are not known and must be estimated from a random sample. Estimation of its parameters has been approached in the literature by various techniques, because a standard maximum likelihood estimate does not exist. To estimate the unknown parameters of the threeparameter inverse Weibull density we will use a combination of onparametric and parametric methods. The idea consists of using two steps: in the first step we calculate an initial nonparametric density estimate which needs to be as good as possible, and in the second step we apply the nonlinear least squares method to estimate the unknown parameters. As a main result, a theorem on the existence of the least squares estimate is obtained, as well as its generalization in the l_p norm (1 p < 1). Some simulations are given to show that our approach is satisfactory if the initial density is of good enough quality.
 D. Jukić, D. Marković, On nonlinear weighted errorsinvariables parameter estimation problem in the threeparameter Weibull model, Applied mathematics and computation 215/10 (2010), 35993609This paper is concerned with the threeparameter Weibull distribution which is widely used as a model in reliability and lifetime studies. In practice, the Weibull model parameters are not known in advance and must be estimated from a random sample. Difficulties in applying the method of maximum likelihood to threeparameter Weibull models have led to a variety of alternative approaches in the literature. In this paper we consider the nonlinear weighted errorsinvariables (EIV) fitting approach. As a main result, two theorems on the existence of the EIV estimate are obtained. An illustrative example is also included.
 D. Jukić, D. Marković, On nonlinear weighted least squares fitting of the threeparameter inverse Weibull distribution, Mathematical Communications 15/1 (2010), 1324In this paper we consider nonlinear least squares fitting of the threeparameter inverse Weibull distribution to the given data (wi; ti; yi), i = 1,...,n, n>3. As the main result, we show that the least squares estimate exists provided that the data satisfy just the following two natural conditions: (i) 0 < t1 < t2 < ... < tn and (ii) 0 < y1 < y2 <... < yn < 1. To this end, an illustrative numerical example is given.
 D. Marković, D. Jukić, On nonlinear weighted total least squares parameter estimation problem for the threeparameter Weibull density, Applied Mathematical Modelling 34/7 (2010), 18391848The threeparameter Weibull density function is widely employed as a model in reliability and lifetime studies. Estimation of its parameters has been approached in the literature by various techniques, because a standard maximum likelihood estimate does not exist. In this paper we consider the nonlinear weighted total least squares fitting approach. As a main result, a theorem on the existence of the total least squares estimate is obtained, as well as its generalization in the total l_q norm ($q\geq 1$). Some numerical simulations to support the theoretical work are given.
 D. Marković, D. Jukić, M. Benšić, Nonlinear weighted least squares estimation of a threeparameter Weibull density with a nonparametric start, Journal of Computational and Applied Mathematics, 228/1 (2009), 304312This paper is concerned with the parameter estimation problem for the threeparameter Weibull density which is widely employed as a model in reliability and lifetime studies. Our approach is a combination of nonparametric and parametric methods. The basic idea is to start with an initial nonparametric density estimate which needs to be as good as possible, and then apply the nonlinear least squares method to estimate the unknown parameters. As a main result, a theorem on the existence of the least squares estimate is obtained. Some simulations are given to show that our approach is satisfactory if the initial density is of good enough quality.
 D. Marković, D. Jukić, A review of some existence results on parameter estimation problem in the threeparameter Weibull model, 12th International Conference on Operational Research, Pula, Croatia, 2008, 103111
 D. Marković, D. Dudaš, Bezierove krivulje i de Casteljauov algoritam, Programski sustav Mathematica u znanosti, tehnologiji i obrazovanju. PrimMath[2003]. , Zagreb, Hrvatska, 2003, 5167Predstavit ćemo neke osnovne ideje CAGDa (Computer Aided Geometric Design). Ključno je da s geometrijskim objektima možemo brzo i učinkovito manipulirati na računalu, tj. da ulazni parametri imaju geometrijsko značenje, drugim riječima da imamo predodžbu kako će zadana krivulja izgledati. U tu svrhu definirat ćemo Bernsteinove polinome, predstaviti njihova osnovna svojstva, te uvesti pojam Bezierovih krivulja kao prikaz polinoma u Bernsteinovoj bazi. Također ćemo pokazati de Casteljauov algoritam. Svi programi bit će izradeni primjenom programskog sustava Mathematica. Pri tome koristit će se grafičke mogućnosti i animacija iterativnog procesa.
 D. Jukić, D. Marković, M. Ribičić Penava, A. Krajina, On the choice of initial approximation of the least squares estimate in some growth models of exponential type, 9th International Conference on Operational Research KOI 2002, Trogir, 2002, 4755
 D. Marković, QR dekompozicija velikih vrpčastih matrica i primjena na konstrukciju aproksimirajućeg splinea, Programski sustav Mathematica u znanosti, tehnologiji i obrazovanju. PrimMath[2001], Zagreb, 2001, 215227
 M. Jukić Bokun, Lj. Jukić Matić, D. Marković, Projekt GAMMA, Osječki matematički list 22/2 (2022), 161170Projekt GAMMA je Erasmus+ projekt na kojemu je Odjel za matematiku Sveučilišta u Osijeku koordinator. U radu predstavljamo projekt i njegove rezultate, opisujemo aktivnosti važne za razvoj rezultata i najavljujemo službene diseminacijske aktivnosti.
Projects

Problem procjene parametara u nekim dvoparametarskim monotonim matematičkim modelima (Parameter estimation problem in some twoparameter monotonic mathematical models)
Scientific project run in 2013/14; supported by the J. J. Strossmayer University of Osijek (principal investigator)
Participation (as researcher) in work of following projects funded by Croatian Ministry of Science:
 Scientific project "Passive control of mechanical models" (23523528181042) within scientific program "Various aspects of parameter estimation problem in nonlinear mathematical models“ (2352818), since 2007;
 Scientific project "Parameter estimation in mathematical models“ (0235001), 2002  2006;
Professional Activities
Refereeing/Reviewing
Osječki matematički list
International Journal of Applied Mathematics and Computer Science (AMCS)
Croatian Operational Research Review (CRORR)
Journal of Risk and Reliability
Conferences and Workshops
with talk:
 On the existence of the nonlinear weighted least squares estimate for some special exponential type models, 15th International Conference on Operational Research, Osijek, September 2014
 On parameter estimation by nonlinear least squares in some special twoparameter exponential type models, International Conference on Advances in Applied Mathematics and Mathematical Physics, Istanbul, Turkey, August 2014
 Parameter estimation problem for Weibull model, 14th Young Statisticians Meeting, Basovizza, October 2009
 Least squares fitting the threeparameter inverse Weibull density, Applied Mathematics and Scientific Computing, Zadar, September 2009
 On nonlinear weighted least squares fitting of the threeparameter inverse Weibull distribution, Applied Mathematics and Scientific Computing, Zadar, September 2009
 A review of some existence results on parameter estimation problem in the threeparameter Weibull model, 12th International Conference on Operational Research, Pula, October 2008
 On nonlinear weighted errorsinvariables parameter estimation problem in the threeparameter Weibull model, 4th Croatian Mathematical Congress, Osijek, June 2008
 Bézierove krivulje i de Casteljauov algoritam, PrimMath[2003], Zagreb, September 2003
 On the choice of initial approximation of the least squares estimate in some growth models of exponential type, 9th International Conference on Operational Research, Trogir, October 2002
 QR dekompozicija velikih vrpčastih matrica i primjena na konstrukciju aproksimirajućeg splinea, PrimMath[2001], Mathematica u znanosti, tehnologiji i obrazovanju, Zagreb, September 2001
without talk:
 5th Croatian Mathematical Congress, Rijeka, June 2012
 4th Conference on Applied Mathematics and Scientic Computing, Brijuni, June 2005
 3rd Conference on Applied Mathematics and Scientic Computing, Brijuni, June 2003
 8th International Conference on Operational Research, Rovinj, September 2000
 2nd Croatian Mathematical Congress, Zagreb, June 2000
Invited Lectures
 Problem procjene parametara u 3parametarskom Weibullovu modelu, Stručnoznanstveni kolokvij AMACIZa,Zagreb, March 2009
 Težinski splineovi, Matematički kolokvij, Osijek, January 2004
 Least squares spline, Matematički kolokvij, Osijek, March 2002
Study Visits Abroad and Professional Improvements
 Eidgenössische Technische Hochschule Zürich, Switzerland, February, 2008
 Technische Universität Berlin, Germany, June 2007
 Max Planck Institut für Informatik, Saarbrücken, Germany, AugustSeptember 2005
Professional Society Membership
 HMD  Croatian Mathematical Society, Department Osijek
 HDOI  Croatian Operational Research Society
Teaching
Konzultacije (Office Hours): konzultacije se održavaju po dogovoru.
Prijedlog tema diplomskih radova na Odjelu za matematiku:
 Polinomijalne matrice;
 Parovi polinomijalnih matrica;
 Osnovna svojstva racionalnih matrica;
 Matrične grupe;
 Egzistencija i konstrukcija generaliziranog inverza
 Moguće su i druge teme u dogovoru sa zainteresiranim studentom
Nastavne aktivnosti u zimskom semestru akademske 2020./2021.
 Geometrija ravnine i prostora  Uvod u algebru, Odjel za fiziku
 Osnove umjetne inteligencije, Odjel za fiziku
 Primijenjena i inženjerska matematika, Prehrambenotehnološki fakultet
Nastavne aktivnosti u ljetnom semestru akademske 2020./2021.
 Metodika nastave informatike 1, Odjel za matematiku
 Metodika nastave informatike, Odjel za matematiku
 Osnove umjetne inteligencije, Odjel za matematiku
Teaching experience
List of courses taught: Linear Algebra 1, Linear Algebra 2, Differential calculus, Integral calculus, Numerical mathematics, Ordinary Differential Equations, Vector Spaces, Algebra, Elementary Mathematics 2, Analytic Geometry, Didactics of Mathematics 2, Mathematics (Faculty of Economics), Applied Mathematics (Faculty of Food Technology), Engineering Mathematics (Faculty of Food Technology), Mathematics 1 (Department of Physics), Mathematics 2 (Department of Physics), Mathematics 3 (Faculty of Electrical Engineering)
Personal
 Birthdate: July 7, 1976
 Birthplace: Osijek, Croatia
 Citizenship: Croatian