Machine Learning
2017/18
Instructors:
Alessandro Verri (alessandro.verri [at] unige.it)
Lorenzo Rosasco (lorenzo.rosasco [at] unige.it)TA:
Fabio Anselmi (anselmi [at] mit.edu) Raffaello Camoriano (raffaello.camoriano [at] iit.it)
Luigi Carratino (luigi.carratino [at] dibris.unige.it)Class Times:
Mon: 11-13am; Wed: 9-11am; Fri: 9-11am; Exceptions: See syllabus
Location:
DIBRIS-room 216; DIBRIS-lab Software 2 (SW2), 3rd floor
Office Hours:
Thu 2-3pm, office 320
Course description
Machine Learning is a key to develop intelligent systems and analyze data in science and engineering. Machine learning engines enable intelligent technologies such as Siri, Kinect or Google self driving car, to name a few. At the same time machine learning methods help unlocking the information in our DNA and make sense of the flood of information gathered on the web, forming the basis of a new Science of Data. This course provides an introduction to the fundamental methods at the core of modern machine learning. It covers theoretical foundations as well as essential algorithms for supervised and unsupervised learning. Classes on theoretical and algorithmic aspects are complemented by practical lab sessions.
More info: http://computerscience.dibris.unige.it/course/ml/
2016/17 edition: mit.lcsl.edu/courses/ml/1617/index.html
Prerequisites
The mathematical tools needed for the course will be covered in some classes in the 2 weeks that precede the course.
Grading
9 CFU course (standard):
- attending classes: 5 points participation, 8 points quiz, 10 points project, 10 points oral discussion
- not attending classes: 10 points project, 23 points oral discussion
< 9 CFU course:
- attending classes: 5 points participation, 8 points quiz, 20 points oral discussion
- not attending classes: 33 points oral discussion
Note 1: The complexity of the oral discussion will be proportional to the number of points.
Note 2: People not attending labs are required to submit one report per lab proving and discussing the results of that lab.
Syllabus
Some of the later classes may be subject to reordering or rescheduling.
Date
Title
Lecturer
Resources
Class 1
Mon 16 Oct
Introduction
AV
Class 2
Wed 18 Oct
Statistical Learning Theory
AV
Class 3
Fri 20 Oct
Lab Octave/Matlab goal: basic matlab/octave+ data generation
AV, FA, LC
Class 4
Mon 23 Oct
Local Methods
AV
Class 5
Wed 25 Oct
Bias Variance Trade-Off
AV
Class 6
Fri 27 Oct
Lab on LM: K-NN, PW for classification
AV, FA, LC
Class 7
Mon 30 Oct
Least Squares Regression
AV
Class 8
Tue 31 Oct
(9:00am - 11:00am
OR 4:00pm - 6:00pm)Least Squares Classification
AV
Class 9
Fri 3 Nov
Lab LS/LDA
LR, FA, LC
Class 10
Mon 6 Nov
Feature Maps
LR
Class 11
Wed 8 Nov
Kernels
LR
Class 12
Fri 10 Nov
Lab Kernels
LR, FA, LC
Class 13
Mon 13 Nov
Regularization Networks and Representer Theorem
LR
Class 14
Wed 15 Nov
Logistic Regression & Support Vector Machines
LR
Class 15
Fri 17 Nov
Lab Loss functions
LR, FA, LC
Class 16
Mon 20 Nov
Dimensionality Reduction
LR
Class 17
Wed 22 Nov
Variable Selection & Sparsity
LR
Class 18
Fri 24 Nov
Lab Dimensionality Reduction and Variable Selection
LR, FA, LC
Class 19
Mon 27 Nov
Density and support estimation
LR
Class 20
Wed 29 Nov
Clustering & K-Means
LR
Class 21
Fri 1 Dec
Lab Clustering
LR, FA, LC
Class 22
Wed 6 Dec
Bayesian ML
AV
Class 23
Mon 11 Dec
Graph Regularization
LR
Class 24
Wed 13 Dec
Multitask Learning
LR
Class 25+26
Thu 14 Dec
(2:00pm - 6:00pm)
Double LabLearning Pipeline - time is (2pm - 6pm)
LR, FA, LC
Class 27
Mon 18 Dec
Neural Networks
LR
References
L. Rosasco. Introductory Machine Learning Notes.
If you have suggestions or find any typos in the notes please fill this form:https://goo.gl/forms/OPJ9Ggk1aWYObYCB3
Hastie, Tibshirani and Friedman. Elements of statistical learning.
Larry Wasserman. Clustering chapter