TIME SERIES ANALYSIS FOR COMPUTER SCIENCE

This is a master's level course on temporal modeling and is part of the interdisciplinary teachings of the master's degree program in Computer Science ("Artificial Intelligence and Data Engineering"). Together with the other courses in statistics and probability
offers a range of modern statistical techniques that are the most in-demand skills these days. The major emphasis is on statistical models for discrete-time data. The focus is on applications with real data and their analysis with statistical programs such as R.

Regular and active participation in the teaching activities offered by the course and in independent research activities will enable students to
1. (Knowledge and understanding)

Create powerful visualizations of temporal data
Carry out exploratory analyses of temporal data
Estimate model parameters using modern statistical software (R)

2. (Knowing and understanding how to apply)
- Select appropriate statistical models for different types of temporal data
- Communicate the analysis effectively in a written paper and an oral presentation.
3. (Judgement)
- autonomously judge the validity and feasibility of different forecasting techniques and understand their impact on the results of analyses

Knowledge of Probability and Statistics at level of Baron M (2019). Probability and Statistics for Computer Scientists. Edition. CRC Press.

Exploring Time Series Data
Statistical Basics for Time Series Analysis
ARMA, ARIMA and Seasonal Models
Time Series Regression
Multivariate Time Series
Deep Neural Network-Based Time Series Models

W. A. Woodward, B. P. Sadler and S.D. Robertson (2022) Time Series for Data Science, CRC Press
R.H. Shumway and D.S. Stoffer (2017) Time Series Analysis and Its Applications: With R Examples. Springer

Additional readings and materials distributed during the course through Moodle

The achievement of the course objectives is assessed through the oral discussion of a project agreed with the teacher.
The project consists in analyzing a data set using the methods learned in the course or, alternatively, Reading a scientific article and reproducing the results with the help of software that was implemented ad hoc in R.
The student is required to prepare a report and then discuss the report with the teacher.
During the oral examination, questions may be asked about parts of the program not covered in the report.

Grades will be determined by oral exam (50%) and by project (50%).

Student will be evaluated in terms of
- quality of his/her statistical analyses
- correct use of the technical terminology
- correct conclusions
- quality of presentation (report)
- quality of the oral discussion

Rules:
1) if the student fails the exam, they can try another session with the *same* project. However, if the exam is failed again, then a *new* project must be considered for the subsequent exam sessions
2) if the student passes the exam but decides to decline the score, then a *new* project must be considered for the subsequent exam sessions

This course is based on lectures, which will cover the major topics, emphasizing and discussing the important points.
Theoretical lectures will be complemented by exercise classes and lab sessions. The statistical software used in the course is R (www.r-project.org).
The personal participation is important, and it will help the student to learn more efficiently to read the assigned material to reinforce the lectures.

English

The attendance to each class (listening) and a genuine effort in doing the suggested exercises are two of the most important factors affecting the success in this course. Statistical analysis require both listening and doing.

oral