Arborescence des pages
ConfigureOutils de l'espace
Aller directement à la fin des métadonnées
Aller au début des métadonnées

GIMAS8AD

Modelling and Forecasting

Duration: 42 hours

ECTS Credits: 4  

Semester: S8

Person(s) in charge:

Sandie FERRIGNO, Associate Professor, Frédéric SUR, Associate Professor, sandie.ferrigno@mines-nancy.univ-lorraine.fr, frederic.sur@mines-nancy.univ-lorraine.fr

Keywords: SAS Software, linear regression, time series, smoothing, Box and Jenkins’ method

Prerequisites: Basic notions of stochastic analysis and statistics: random variables, estimators, statistical hypothesis testing

Objective:

forecast a phenomenon, using a model based on its past and / or its context

Program and contents:

Introduction to SAS software

Regression as a modelling tool:

  • simple and multiple linear regression models
  • regression control (local and global quality indexes)

  • regression models selection (Mallows’ Cp, step by step procedures)

  • regression validation (analysis of residuals, influential or suspicious observations)

     

Time series analysis and forecasting:

  • time series decomposition: trend and seasonal structure analysis.
  • Smoothing techniques

  • ARMA, ARIMA, SARIMA models,  Box and Jenkins’ method.

  • multivariate time series and intervention models.

 

Abilities: 

Levels

Description and operational verbs

Know 

 

Understand 

 

Apply 

 

Analyze 

 

Summarise

 

Assess

 

Evaluation:

  • Written test
  • Continuous Control
  • Oral report
  • Project
  • Written report
  • Aucune étiquette