Pattern Analysis and Machine Intelligence
The following are last minute news you should be aware of ;-)
10/10/2014: The new course edition is about to start!!! Stay tuned it will be significantly different from the past (check the 'new words in the text)!!!!
Course Aim & Organization
The objective of this course is to give an advanced presentation, i.e., a statistical perspective, of the techniques most used in artificial intelligence and machine learning for pattern recognition, knowledge discovery, and data analysis/modeling.
Teachers
The course is composed by a blending of lectures and exercises by the course teacher and a teaching assistant.
- Matteo Matteucci: the course teacher
- Davide Eynard: the teaching assistant
Course Program
Techniques from machine and statistical learning are presented from a theoretical (i.e., statistics and information theory) and practical perspective through the descriptions of algorithms, the theory behind them, their implementation issues, and few examples from real applications. The course follows, at least partially, the following new book which is available for download in pdf
- An Introduction to Statistical Learning with Applications in R by Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani
A tentative syllabus for this year edition is the following (a detailed schedule of the lectures follows)
- Machine Learning and Pattern Classification: in this part of the course the general concepts of Machine Learning and Patter Recognition are introduced with a brief review of statistics and information theory;
- Unsupervised Learning Techniques: the most common approaches to unsupervised learning are described mostly focusing on clustering techniques, rule induction, Bayesian networks and density estimators using mixure models;
- Supervised Learning Techniques: in this part of the course the most common techniques for Supervised Learning are described: decision trees, decision rules, Bayesian classifiers, hidden markov models, lazy learners, etc.
- Feature Selection and Reduction: techniques for data rediction and feature selection will be presented with theory and applications
- Model Validation and Selection: model validation and selection are orthogonal issues to previous technique; during the course the fundamentals are described and discussed (e.g., AIC, BIC, cross-validation, etc. ).
Detailed course schedule
A detailed schedule of the course can be found here; topics are just indicative while days and teachers are correct up to some last minute change (I will notify you by email). Please note that not all days we have lectures!!
Note: Lecture timetable interpretation * On Mondays lectures start at 13:15 SHARP! * On Fridays lectures start at 10:30 (quarto d'ora accademico)
(*) With respect to following version of the book PDF file of book (10th printing with corrections, Jan 2013)
- Chapter 2 is a good introductory chapter.
- in particular refer to 2.3.1 for an introduction to least squares
- section 2.4 for an introduction to statistical decision theory
- section 2.5 for an example and 2.9 for a discussion about bias and variance decomposition of prediction error
- section 2.8 for introduction to restricted (penalized) estimators
- Chapter 3 is the most important chapter for linear regression
- 3.1
- 3.2 (you can skip confidence internals for \beta, but study the Z-scores)
- 3.2.2 the proof is not requires, the statement of the theorem is important
- 3.2.3 you can skip the regression by successive orthogonalizations, however dont skip the first part of the section
- 3.2.4 skip this section
- 3.3 very important
- 3.4 very important (i will not ask the formulas for the degrees of freedom of ridge and lasso)
- 3.5 you can skip this
- 3.6 take a look
- 3.7 you can skip this
- 3.8 take a look, you can skip 3.8.3 to 3.8.6
- 3.9 take a look
- As to the paper Least Angle Regression take a look at section 1 to 3
Course Evaluation
The new course evaluation is composed by two parts:
- HW: Homework with exercises covering the whole program
- WE: A written examination covering the whole program
the final score will take the MAXIMUM between WE' and the combination 0.7*WE + 0.3*HW. In practice
- the homework can only increase your score
- the homework can only impact for the 30% of the score
- the homework is not mandatory
Teaching Material (the textbook)
Lectures will be based on material taken from the book.
- An Introduction to Statistical Learning with Applications in R by Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani
If you are interested in a more deep treatment of the topics you can refer to the following book from the same authors
- The Elements of Statistical Learning: Data Mining, Inference, and Prediction. by Trevor Hastie, Robert Tibshirani, and Jerome Friedman.
Some additional material that could be used to prepare the oral examination will be provided together with the past homeworks.
Teacher Slides
In the following you can find the lecture slides used by the teacher and the teaching assistants during classes:
- Support Vector Machines: these slides have been used to present Support Vector Machines (taken from The Elements of Statistical Learning book).
-->
Additional Papers
Papers used to integrate the textbook
Past Exams and Sample Questions
This edition of the course in new so the exams of this year will vary in the format
- 2 theoretical questions
- 2 practical exercises
These are the text of past exams to give and idea on what to expect a theoretical questions:
- 20/09/2013 Exam
- 10/09/2013 Exam
- 26/07/2013 Exam
- 11/07/2013 Exam
- 29/01/2013 Exam
- 19/09/2012 Exam
- 04/09/2012 Exam
- 10/07/2012 Exam
- 26/06/2012 Exam
- 03/02/2012 Exam
- 19/09/2011 Exam
- 08/09/2011 Exam
- 15/07/2011 Exam
- 29/06/2011 Exam
Online Resources
The following are links to online sources which might be useful to complement the material above
- MATH 574M University of Arizona Course on Statistical Machine Learning and Data Mining; here you can find slides covering part of the course topics (the reference book for this course is again The Elements of Statistical Learning)
2013-2014 Homework
The 2013 Homework (alike the 2012 one) is organized as an octave series of tutorials. You are requested to go through the tutorials and practice with the algorithms that have been presented in class. To prove us you have done it and that you have understood the code you will be requested to solve few exercises and provide us a pdf report by email
Part 1: Linear Classification Methods
- Homework 2013-2014 on Classification: this is the text of the first part of the homework; it has been intentionally edited not to allow cut and paste. This part of the homework will contribute to the 10% of the grade and the deadline to submit the solution by Sunday 17/11 23:59
- SAheart.data: the dataset used for the homework
- SAheart.info: the dataset used for the homework
Note 1: Submit the solution by loading it on www.dropitto.me/matteucci (pwd is dropittome)
Note 2: please name your pdf as pami_SURNAME_STUDENTID_classification.pdf; if you submit a homework for different people, please pick one of the names for the file but PUT ALL THE NAMES IN THE COVER PAGE!!
Part 2: Regression
- Homework 2013-2014 Regression: this is the text of the second part of the homework; it has been intentionally edited not to allow cut and paste. This part of the homework will contribute to the 10% of the grade and the deadline to submit the solution by email to malago@di.unimi.it (cc to matteo.matteucci@polimi.it) is Friday 20/12 23:59
- prostate.data: the dataset used for the homework
- prostate.info: the dataset used for the homework
- diabete.mat: the dataset used for the homework
- textread.m: (optional) function which might be useful depending on your octave version
- strread.m: (optional) function which might be useful depending on your octave version
For any question or doubt please sen us an email as soon as possible.
Note 1: for some strange reason the CSM of the website has decided to rename the files with capitals, please save them in lower case :-(
Note 2: rename the file Diabete.data into diabete.mat ... still fighting with the CSM :-)
Note 3: the code has been tested with octave under linux, we suggest to use it not to spend too much time with installing it under windows or using matlab. If you do not have linux installed, try using a live CD as the ubuntu 13.04 live distro ;-)
Part 3: Clustering
The code and the text of the third part of the homework are available online at this post
As usual, this part of the homework will contribute to the 10% of the grade; the deadline to submit the solution is the end of the course. You have to send it to davide.eynard_at_gmail.com Friday 24/01 23:59.
Note 1: for any doubt or question send an email, as soon as possible, to Davide Eynard so to have a prompt reply and not get stuck during homework execution.
Note 2: you have to turn in only the solution of "Ocatave clustering demo part 6", while the other parts can be used as reference to improve your understanding in basic clustering algorithms.