Forecasting Methods and Time Series

Code:

2MADSAD09

Acronym:

MPST

Keywords
Classification	Keyword
OFICIAL	Statistics

Instance: 2015/2016 - 2S

Active?	Yes
Responsible unit:	Agrupamento Científico de Matemática e Sistemas de Informação
Course/CS Responsible:	Master in Modeling, Data Analysis and Decision Support Systems

Cycles of Study/Courses

Acronym	No. of Students	Study Plan	Curricular Years	Credits UCN	Credits ECTS	Contact hours	Total Time
MADSAD	14	Bologna Official Syllabus	1	-	7,5	56	202,5

Teaching language

English

Objectives

The aim of this course is to introduce the students to time series analysis methods.

Learning outcomes and competences

By the end of the course, the student should:

1. understand basic time series concepts and terminology
2. be able to select time series methods appropriate to forecast
3. be able to use apropriate software
4. be able to concisely summarize results of a time series analysis

Working method

Presencial

Pre-requirements (prior knowledge) and co-requirements (common knowledge)

Introductory Statistics

Program

1. Introduction: definition of time series, the aims of time series analysis, examples of time series. Descriptive analysis: cronogram, identification and removal of trend and seasonal components, transformations. Fundamentals of stochastic processes: definition; stationarity; weak stationarity; autocovariance and autocorrelation
functions, partial autocorrelation function; linear difference equations. Estimating the mean and the
autocovariance and autocorrelation functions. Measuring the precision of predtions.

2. Exponential smoothing methods. Moving averages. Simple exponential smoothing. Double exponential smoothing. Triple exponential smoothing.

3. Time series decomposition. Decomposition models: additive and multiplicative; Loess; "bureau of census"; STL. Prediction.

4. Stationary linear time series models: autorregressive models (AR), moving average models (MA), ARMA and SARMA models. Models for linear non stationary time series: ARIMA and SARIMA models. Prediction. Box-Jenkins approach to time series analysis: identification, estimation, model checking. Regression and time series.

5. Unit roots tests
5.1 Dickey-Fuller and Augmented Dickey-Fuller tests.
5.2 Phillips-Perron tests.
5.3 KPSS test.

Mandatory literature

Box, G.E.P., Jenkins, G.M. e Reinsel, G.C. ; Time Series Analysis: Forecasting and Control. 3ª ed., John Wiley & Sons, 1993
Makridakis, S., Wheelright, S. e Hyndman, R. ; Forecasting: Methods and Applications. 2ª ed., John Wiley and Sons , 1996
Brockwell, P.J. e Davis, R.A. ; ntroduction to Time Series and Forecasting, Springer-Verlag, 1996
Hamilton, J.; Time Series Analysis., Princeton University Press , 1994
Lutkepohl, H. e Kratzig, M. ; Applied Time Series Econometrics., Cambridge University Press , 2004
Wei, William W. S.; Time series analysis. ISBN: 0-201-15911-2

Teaching methods and learning activities

Classes; example classes and laboratory classes.

Software

R project
R project

keywords

Physical sciences > Mathematics > Statistics

Evaluation Type

Distributed evaluation with final exam

Assessment Components

Designation	Weight (%)
Exame	50,00
Trabalho escrito	50,00
Total:	100,00

Calculation formula of final grade

Exam 50% + Project 50%.

Minimum grade of 7/20 for each of the components. If a student obtains a grade less than 7/20 (the minimum), his final grade will be 7.

Classification improvement

The student must take the exam and resubmit a project.