Demand and Sales Forecasting

Goals: 

Concepts for managing an archive of historical data, carrying out operations of selection, transformation, aggregation, etc.

​

Techniques presented: 

Preliminary analysis of the historical series, through graphical representations and study of the autocorrelations.

​

Tutorials:

​There are exercises for each of the topics covered.

​

Prerequisites: 

Attendance at the INTRO course or having an intermediate knowledge of Statistics for Data Analysis is a prerequisite.

​

Subjects:

The linear regression

• Recall of simple linear regression

• multiple linear regression

  • Hypothesis testing on parameters

  • automatic selection of variables

  • Residual analysis

  • Multicollinearity

  • Using the model for prediction and simulation

• Introduction to the generalized linear model

  • The analysis of variance

  • Mixed effects models

  • Multivariate analysis of variance and repeated measures

 

Introduction to time series analysis

Definition of time series

What is a good forecast

Conceptual scheme of a forecasting system

Choosing a model

Forecasting analysis applications

• Useful operations and features of Statistics for Data Analysis

• Regression models and their applications

 

Econometric models: econometrics and regression

The decompositional approach

• Multiplicative and additive model in the treatment of seasonal components

 

Exponential leveling

• Forecasting with levelings

 

Stochastic processes

• ARIMA models

  • Building a  seasonal arima model

• Intervention analysis: elements related to ARIMA models

  • Prediction and simulation with ARIMA models using dummy variables

  • Treatment of structural breaks with ARIMA models

• ARMAX models: use of covariates

• Multivariate models: standard VAR models

 

Performance of forecasting models, introduction to simulation and automatic forecasting.