POLITECNICO DI TORINO - DIPARTIMENTO DI MATEMATICA - GRUPPO DI STATISTICA E PROBABILITA'

Venerd́ 11 Marzo 2011	ore 14:00	Henry Laniado-Rodas Universidad Carlos III de Madrid	On the Portfolio Selection Problem The portfolio selection problem was considered in Markowitz, where the philosophy is that an investor should hold a portfolio on the set of risk-return couples that cannot be improved simultaneously. This set in R2 is denoted as the efficient frontier. Several criteria have been studied for portfolio selection in Markowitz. In this talk will be introduced what is the specific problem and will be discussed different methodologies of solution.
Luned́ 21 Marzo 2011	ore 16:30	Mahdi Roozbeh Ferdowsi University of Mashhad	Estimation of Partial Linear Model with Differencing Methodology In this paper, we will estimate the parametric and nonparametric part of a partial linear model by differencing method under multicollinearity and non-stochastic linear hypothesis settings. Partial linear models are more flexible than standard linear models since they have a parametric and a nonparametric component. Multicollinearity is defined as the existence of nearly linear dependency among column vectors of the design matrix. The idea of differencing to remove the nonparametric effect in nonparametric regression models is not new. In the partial linear model, it has been examined by Ahn and Powell (1993). In a pure nonparametric regression setting, the idea of differencing has a longer history and has been used to obtain estimators of the residual variance. The difference-based estimation procedure is optimal in the sense that the estimator of the linear component is asymptotically efficient and the estimator of the nonparametric component is asymptotically minimax rate optimal for the partial linear model (Wang et al. 2007). In this approach, we use differencing method to remove the nonparametric part of partial linear model in order to estimate the linear part. After estimation of the linear part, a variety of nonparametric techniques can be applied to estimate the nonparametric part.
Marted́ 19 Aprile 2011	ore 13:30 aula 7D	Mauro Gasparini Politecnico di Torino	Dose-finding problem as an inverse regression problem Dose-finding activities at the beginning of Clinical Development of a new therapy present several peculiarities: a high degree of adaptivity, the need to provide safe (from below) dose escalation, the discreteness of the grid of possible dose values and the opportunity to reach decisions in the presence of small sample sizes. These aspects have given medical and statistical researchers the opportunity to discuss new statistical methodologies and to compare several different approaches which will be introduced and discussed. References are Ting, N., ed. (2006). Dose Finding in Drug Development. Springer, Chevret, S., ed. (2006), Statistical Methods for Dose-Finding Experiments, Wiley, and a recent issue of Statistical Science (May 2010), containing five articles devoted to the subject. Extensions to two dimensions are currently studied and will be discussed, in particular the relatioships with the literature on copulas and binary multiple regression.
Mer. 20 Aprile 2011	ore 15:00 aula 7D	Roberto Fontana Politecnico di Torino	Algebraic generation of orthogonal fractional factorial designs: some results based on integer linear programming Generation of orthogonal fractional factorial designs (OFFD) is a relevant and extensively studied subject in applied statistics. In this paper we show how searching for an OFFD that satisfies a set of constraints, expressed in terms of orthogonality between simple and interaction effects, is, in many applications, equivalent to solving an integer linear programming problem. We use a recent methodology (Fontana and Pistone, 2010), based on polynomial counting functions and strata, that represents OFFDs as the positive integer solutions of a system of linear equations . Then we set up an optimization problem where the cost function to be minimized is, for example, the size of the OFFD and the constraints are represented by the system. Finally we search for a solution using standard integer programming techniques. Some applications are presented. It is worth noting that the methodology does not put any restriction on the number of levels of each factor and so it can be applied to a really wide range of designs, including mixed orthogonal arrays. The work has been presented at the ENBIS DEINDE 2011 Spring Conference, Turin, March 17, 2011.
Luned́ 16 Maggio 2011	ore 10:00 aula Buzano	Benjamin Ribba INRIA Rhône-Alpes, Lyon	Modeling tumor growth and response to treatments in the context of drug development I
Marted́ 17 Maggio 2011	ore 13:00 aula Buzano	Benjamin Ribba INRIA Rhône-Alpes, Lyon	Modeling tumor growth and response to treatments in the context of drug development II
Luned́ 13 Giugno 2011	ore 13:00 aula Buzano	Domenico Piccolo Università di Napoli Federico II	Una modellistica per dati ordinali: inferenza statistica ed esperienze reali
Luned́ 26 Settembre 2011	ore 10:00 aula Buzano		Workshop: Modelli Statistici per la Customer Satisfaction