Applied Forecasting

The classes are, according to what is determined in the curricular plan the expository method will be used to introduceconcepts, fundamental results and methodologies. The practice sessions will be aimed at exemplifying procedures andproblem solving with the support of a computer tool (predominantly, R or Python). A strong interaction between theoryand practice will prevail, giving, as much as possible, a central role to the visualization and treatment of concrete
situations.
The student will be graded by a project made within the course and by a written exam. The written exam is mandatory,while the project is optional. The final grade is equal to 50% of the project grade plus 50% of the written exam result.The project grade will only be considered if the student obtains a minimum result of 7 in the written exam (on a scale of0 to 20). If the student chooses not to carry out the project, the final grade will be entirely determined by the writtenexam.

Data description and forecasting tasks often involve data where the time component cannot be neglected.
Thus, in this curricular unit, the focus is on time series data. It is intended that the student is able to:
− carry out the exploratory analysis adjusted to the time series of interest;
− perform the decomposition of the series into its components;
− identify potential problems that interfere with the quality of the data and carry out the due treatment;
− recognize different classes of forecasting methods for time series;
− identify and apply the appropriate forecast method(s) in specific cases;
− know the assumptions and limitations of the methods;
− know how to validate the obtained models and evaluate their predictive capacity.
At the same time, the student must be able to perform the tasks in an efficient manner using the appropriatecomputer(s) tool, preferably, R or Python.

1. Time series description and pre-processing
1.1. Exploratory analysis
1.2. Decomposition
1.3. Pre-processing
2. Time series forecasting
2.1. Toolkit: concepts, procedures and measures
2.2. Exponential smoothing
2.3. ARIMA models
2.4. Dynamic regression models
2.5. Advanced forecasting methods

NAO

Box, G.E., Jenkins, G.M., Reinsel G.C. (2015). Time Series Analysis: Forecasting and Control, 4th edition. Wiley.
Gilliland, M., Tashman, L., Sglavo, U. (2015). Business Forecasting: Practical Problems and Solutions. Wiley.
Hyndman, R.J., Athanasopoulos, G. (2021). Forecasting: Principles and Practice, 3rd edition. OTexts.https://otexts.com/fpp3/
Krispin, R. (2019). Hands-On Time Series Analysis with R. Packt Publishing.
Lazzeri, F. (2020). Machine Learning for Time Series Forecasting with Python. Wiley.
Nielsen, A. (2019). Practical Time Series Analysis: Prediction with Statistics and Machine Learning. O’Reilly Media.