2024
Forecasting Models
Name: Forecasting Models
Code: MAT13637L
6 ECTS
Duration: 15 weeks/156 hours
Scientific Area:
Mathematics
Teaching languages: Portuguese
Languages of tutoring support: Portuguese
Sustainable Development Goals
Learning Goals
To study models that incorporate the influence of one or several explanatory variables in a certain variable of interest. Linear models in which the assumption of the independence of the response variable is not met, time series models of the SARIMA type, extensions to linear regression models are studied for discrete and continuous data, in which the assumption of normality of the distribution of errors is not met.
Competences:
To acquire fundamental principles in statistical modelling and forecasting, with particular emphasis on application to economics, finance and management data.
To develop in students the ability to, in a critically and autonomously way, interpret problems, identify the appropriate models to use according to the nature of the response variable (discrete or continuous), whenever the assumptions of the usual models are not met, and apply these methodologies in their professional activity.
To correctly use appropriate statistical software.
Competences:
To acquire fundamental principles in statistical modelling and forecasting, with particular emphasis on application to economics, finance and management data.
To develop in students the ability to, in a critically and autonomously way, interpret problems, identify the appropriate models to use according to the nature of the response variable (discrete or continuous), whenever the assumptions of the usual models are not met, and apply these methodologies in their professional activity.
To correctly use appropriate statistical software.
Contents
Review of general concepts of stochastic processes:
Stationary and non-stationary stochastic processes.
Basic concepts of time series.
Patterns identification.
Linear models for time series:
ARMA, ARIMA and SARIMA models.
Identification of the models.
Model adjustment.
Forecast.
Introduction to generalised linear models:
Models for counting data.
Models for continuous data.
Model adjustment.
Forecast.
Application to economic and financial data using statistical software.
Stationary and non-stationary stochastic processes.
Basic concepts of time series.
Patterns identification.
Linear models for time series:
ARMA, ARIMA and SARIMA models.
Identification of the models.
Model adjustment.
Forecast.
Introduction to generalised linear models:
Models for counting data.
Models for continuous data.
Model adjustment.
Forecast.
Application to economic and financial data using statistical software.
Teaching Methods
The classes are theoretical-practices in the blackboard, with support of e-learning tools and slides, and with the use of the available statistical software for the exercises. To motivate students attendance to the classroom and students continuous work.
Introduction of theoretical concepts using examples of direct application in different areas to illustrate the importance of course contents. Exercises with emphasis in the resolution of real problems, to stimulate interest in the course and to demonstrate its utility.
Introduction of theoretical concepts using examples of direct application in different areas to illustrate the importance of course contents. Exercises with emphasis in the resolution of real problems, to stimulate interest in the course and to demonstrate its utility.
Assessment
To privilege continuous evaluation, with tests (50%) and individual or group homeworks (50%).
Evaluation under examination: a final exam (50%), where it will be necessary to use the computer to answer the questions, and an application homework (50%).
Evaluation under examination: a final exam (50%), where it will be necessary to use the computer to answer the questions, and an application homework (50%).
