
Course unit
MACHINE LEARNING
SC01122905, A.A. 2017/18
Information concerning the students who enrolled in A.Y. 2017/18
Mutuating
Course unit code 
Course unit name 
Teacher in charge 
Degree course code 
SC01122905 
MACHINE LEARNING 
FABIO AIOLLI 
SC1176 
ECTS: details
Type 
ScientificDisciplinary Sector 
Credits allocated 
Educational activities in elective or integrative disciplines 
INF/01 
Computer Science 
6.0 
Course unit organization
Period 
First semester 
Year 
1st Year 
Teaching method 
frontal 
Type of hours 
Credits 
Teaching hours 
Hours of Individual study 
Shifts 
Laboratory 
1.0 
8 
17.0 
No turn 
Lecture 
5.0 
40 
85.0 
No turn 
Examination board
Board 
From 
To 
Members of the board 
9 a.a. 2019/2020 
01/10/2019 
28/02/2021 
AIOLLI
FABIO
(Presidente)
SPERDUTI
ALESSANDRO
(Membro Effettivo)
BALDAN
PAOLO
(Supplente)
BALLAN
LAMBERTO
(Supplente)
PALAZZI
CLAUDIO ENRICO
(Supplente)

8 a.a. 2018/2019 
01/10/2018 
28/02/2020 
AIOLLI
FABIO
(Presidente)
PALAZZI
CLAUDIO ENRICO
(Membro Effettivo)
BALDAN
PAOLO
(Supplente)
SPERDUTI
ALESSANDRO
(Supplente)
TOLOMEI
GABRIELE
(Supplente)

7 a.a. 2017/2018 
01/10/2017 
28/02/2019 
AIOLLI
FABIO
(Presidente)
BALDAN
PAOLO
(Membro Effettivo)
PALAZZI
CLAUDIO ENRICO
(Membro Effettivo)
SPERDUTI
ALESSANDRO
(Membro Effettivo)
TOLOMEI
GABRIELE
(Membro Effettivo)

Prerequisites:

The student should be familiar with basic concepts in Probability and Analysis of multivariate functions. It is also advisable to have basic knowledge of Programming and Artificial Intelligence.
The course does not have prerequisites. 
Target skills and knowledge:

The aim of the course is to introduce the student to the basic concepts that characterize Machine Learning, i.e. the class of techniques and algorithms that starting from empirical data allow a computer system to aquire new knowledge, or to correct and/or to refine knowledge already available. These techniques are particularly useful for problems for which it is impossible or very difficult to reach a mathematical formalization usable for the definition of an ad hoc algorithmic solution. Examples of these problems are perceptual tasks, such as visual recognition of handwritten digits, or problems in which data is corrupted by noise or is incomplete. The course mainly covers numerical methods.
Students will face practical exercises in a computer lab that allow them to test the application of the acquired knowledge to small practical examples. 
Examination methods:

The student must pass a written examination and, if deemed necessary by the teacher, an oral examination. 
Assessment criteria:

The text of the written exam includes some questions that aim to assess the level of learning reached by the student concerning the concepts taught in the course and the student's ability to perform critical analysis on them. The text also includes questions in which the student is required to show understanding of the applicative issues taught in the computer lab. These questions are designed to assess whether the student has developed the ability to apply the concepts learned during the course.
In the event that the assessment of the written exam is not satisfactory for the student, the teacher may supplement the written examination with an oral examination to better assess the level of learning of the student. 
Course unit contents:

The course will cover the topics listed below
 Introduction:
When to apply Machine Learning techniques; Machine Learning Paradigms; Basic ingredients of Machine Learning.
 Learning Concepts:
Complexity of the Hypothesis Space; Complexity Measures; Examples of Supervised Learning Algorithms;
 Decision Trees:
Learning Decision Trees; Treatment of Numerical Data, Missing Data, Costs; Pruning Techniques and Derivation of Decision Rules.
 Probabilistic Learning:
Bayesian Learning; Examples of Application to Supervised and Unsupervised Learning (clustering); Optimal Bayes classifier; EM.
 Neural Networks and Support Vector Machines:
Introduction to Neural Networks; Classification Margin, Support Vector Machines for Classification and Regression, Kernel Functions.
 Application Issues:
Classification Pipeline, Representation and Selection of Categorical Variables; Model Selection, Holdout, CrossValidation, LeaveOneOut CV; Exterior and Interior Criteria for Evaluating a Clustering System; Recommendation Systems: Types, Approaches, Evaluation Measures. 
Planned learning activities and teaching methods:

The course consists of lectures and exercises in the computer lab. The exercises in the computer lab allow the students to experiment, under various operating scenarios, with the techniques introduced in class. In this way, students can verify experimentally the concepts learned in class and acquire the ability to apply the learned concepts and to perform critical judgment. 
Additional notes about suggested reading:

Slides presented during the lectures are made available to students as reference material. 
Textbooks (and optional supplementary readings) 

Mitchell, Tom M., Machine learning. New York: McGrawHill, 1998.

Alpaydin, Ethem, Introduction to machine learning. Cambridge: The MIT press, 2010.


