Artificial Intelligence for Time Series Analysis
A.Y. 2025/2026
Learning objectives
The course aims to provide theoretical and practical skills for the automatic analysis of time series using statistical and artificial intelligence techniques.
Expected learning outcomes
Students will be able to i) model time series with a probabilistic approach, ii) design time series predictors, and iii) implement statistical and AI models in this context.
Lesson period: Third four month period
Assessment methods: Esame
Assessment result: voto verbalizzato in trentesimi
Single course
This course cannot be attended as a single course. Please check our list of single courses to find the ones available for enrolment.
Course syllabus and organization
Single session
Responsible
Lesson period
Third four month period
Course syllabus
Introduction to the course:
· Structure of the course
· Motivation for the course
· Examples of applications
Stochastic processes:
· Basics of probability space
· Definition and properties of stochastic processes
· Introduction to linear stationary and non-stationary stochastic models
· Seasonality and trend
· Wold decomposition theorem
· Forecasting and smoothing
Maximum likelihood estimation (MLE) and maximum a posteriori (MAP):
· Introduction to MLE and MAP estimation methods
· Adaptation of the method for time series
· Underfitting vs overfitting for time series
Kalman filter and Hidden Markov Model:
· Introduction to state-space models and applications
· Introduction to the Kalman filter and applications
· Introduction to the Markov Process and applications
· Introduction to the Hidden Markov Model and applications
Neural networks for time series:
· Use of convolutional neural networks and skip connection for time series
· Introduction to recurrent neural networks and state-space model
· Examples of applications
· Structure of the course
· Motivation for the course
· Examples of applications
Stochastic processes:
· Basics of probability space
· Definition and properties of stochastic processes
· Introduction to linear stationary and non-stationary stochastic models
· Seasonality and trend
· Wold decomposition theorem
· Forecasting and smoothing
Maximum likelihood estimation (MLE) and maximum a posteriori (MAP):
· Introduction to MLE and MAP estimation methods
· Adaptation of the method for time series
· Underfitting vs overfitting for time series
Kalman filter and Hidden Markov Model:
· Introduction to state-space models and applications
· Introduction to the Kalman filter and applications
· Introduction to the Markov Process and applications
· Introduction to the Hidden Markov Model and applications
Neural networks for time series:
· Use of convolutional neural networks and skip connection for time series
· Introduction to recurrent neural networks and state-space model
· Examples of applications
Prerequisites for admission
Basic knowledge of statistics and probability (e.g. probability distribution, expected value, sample mean, etc.) and discrete mathematics (e.g. vectors, matrices, scalar product, etc.) are recommended.
Teaching methods
The course is structured in lectures of theory and exercises. Some lessons will be enriched by scientific seminars held by experts (based on their availability).
Teaching Resources
The material will be provided entirely by the professor in the form of slides, notes and code on myAriel.
Assessment methods and Criteria
The exam comprises a written exam on the topics of the course and a project on a specific topic selected by the student which will presented with an oral exam. The written exam weights 1/3 of the final grade.
Professor(s)