Statistical and Mathematical Methods for Machine Learning (SMM)

What follows are the teaching materials for the Statistical and Mathematical Methods for Machine Learning (SMM, in the following), Module 2, for the Academic Year 2024-2025, held by professor Davide Evangelista.

Topics

Module 2 of the SMM course is divided into 3 parts. In the first part (which will be covered in the first 2 to 3 lessons), we will briefly recall the basics of Python programming language, which will be our tool to test the algorithms. Moreover, we will introduce numpy (arguably the most used Python library for vector and matrix manipulation), matplotlib (for visualization) and other add-on libraries for numpy such as pandas and scipy.

In the second part of the course, we will introduce the basics of Machine Learning, and in particular we will study how to implement some simple yet effective Machine Learning algorithms for Dimensionality Reduction and Classification. The focus will always be to deeply understand the math behind those concepts, and how their implementation relates to basic linear algebra concepts you already studied during your bachelor degree.

In the last part, we will discuss Optimization techiniques such as Gradient Descent and Stochastic Gradient Descent (SGD), for the solution of the training problem in Machine Learning algorithms. Moreover, we will introduce two frameworks (i.e. Maximum Likelihood Estimation (MLE) and Maximum A Posteriori (MAP)) commonly used in Machine Learning applications, with the aim of understanding the origin of some commonly-used formula you will see in the following courses, over the Master degree.

Tutor of the Course

For this course a tutor is available. In case you need any assistance, please refer to Daniele Gregori (daniele.gregori6@unibo.it).

Office hour

The office hour is Wednesday morning, from 9a.m. to 12p.m., via Microsoft Teams or in my office, at Ex Veneta, close to the Zanolini train station.

Either cases, it is mandatory to set an appointment by e-mail, via the address davide.evangelista5@unibo.it.