The 22nd International Conference on Computational Statistics took place at the Auditorium/Congress Palace Principe Felipe, Oviedo, Spain, 23-26 August 2016. The conference is sponsored by the European Regional Section of the IASC and is organized by the University of Oviedo. It aims at bringing together researchers and practitioners to discuss recent developments in computational methods, methodology for data analysis and applications in statistics.

yuimaGUI: A graphical user interface for computational finance based on the yuima R package

Abstract. The aim of Yuima project is to develop a complete environment for simulation and inference of Stochastic Differential Equations (SDE) via an R package called yuima. The package is developed using the object oriented programming language S4 and allows the user to manage a stochastic process characterized by a SDE with the following general form: dX_t =b(t,X_t,\theta)dt+a(t,X_t,\theta)dW^H_t +c(t,X_t,\theta)dZ_t, where b(t,X_t,\theta), a(t,X_t,\theta) and c(t,X_t,\theta) are functions defined by the user. W^H is the fractional Brownian motion where the Hurst coefficient H is fixed by default to 1/2 that corresponds to the standard Brownian motion. The yuima package has also estimation and simulation routines for two models widely used in modern computational finance: the Continuous ARMA(p,q) driven by a general Levy process and the COGARCH(p,q) model for high frequency data. Moreover, yuimaGUI allows to simulate a portfolio of assets and derivatives under different scenarios and further evaluate thee overall returns along with their distribution. After a brief presentation of the yuima framework, we will focus on aspects related to computational finance and show how they can be easily approached through the yuimaGUI package. We will go through data mining and clustering of financial time series, model selection, change point estimation, scenario simulation and the analysis of the distribution of the expected returns for composite portfolios.

Speaker. Emanuele Guidotti

Authors. Emanuele Guidotti, Stefano Iacus, Lorenzo Mercuri