Dissertation defense (February 19, 2025): Arthur Ronald Ferreira Diogenes Garcia

Student: Arthur Ronald Ferreira Diogenes Garcia

Title: Sliding Window-based Autoregressive Moving-Average

Advisors: Eduardo Soares Ogasawara and Dayse Haime Pastore

Committee: Eduardo Soares Ogasawara (CEFET/RJ), Dayse Haime Pastore (CEFET/RJ),  Jorge de Abreu Soares (CEFET/RJ), Fábio André Machado Porto (LNCC)

Day/Time:  February 19, 2025 / 9 a.m.

Room: https://teams.microsoft.com/l/meetup-join/19%3aKcszJdyzz4_vwVbrPDe0h62eXY262TcHQZS6U1eaeio1%40thread.tacv2/1738630105311?context=%7b%22Tid%22%3a%228eeca404-a47d-4555-a2d4-0f3619041c9c%22%2c%22Oid%22%3a%220534c83a-4cef-4129-9f9f-e8b8c3d50019%22%7d

Abstract: Time series models have been proposed for decades, and most of them require the stationary condition as a prerequisite, i.e., the mean, variance, and covariance of the serie don’t change over time. As a consequence, it is necessary to apply preprocessing methods in order to obtain a stationary serie. However, it is desirable for issues of analyzing that the model as well as the statistical properties of the series can be interpretable in order to assist the user analysis. This paper proposes a model which combines a preprocessing phase that achieves stationarity by preserving the statistical properties of the original serie, and a modelling phase by using Autoregressive Moving Average.