Difference between revisions of "Machine Learning"

Revision as of 12:05, 25 February 2017

The following are last minute news you should be aware of ;-)

* 25/02/2017: Detailed grades for the 13/02/2017 exam are here
* 03/02/2017: The complete homework is out!!!
* 21/01/2016: The homework for this year is out!!
* 15/11/2016: Change to the lecture schedule; several dates, please check!!!
* 08/10/2016: Updated all course slides
* 23/10/2016: Change in the detailed schedule for 8/11, 15/11, 2/12, 6/12 ... just a couple of swaps
* 04/10/2016: Lectures start today!
* 01/09/2016: This is the Machine Learning course website. The course is basically a rename of the PAMI course.

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. The course will provide the basics of Regression, Classification, and Clustering with practical exercises using the R language.

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., based on 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 mostly follows the following book which is also available for download in pdf

• An Introduction to Statistical Learning with Applications in R by Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani

The course is composed by a set of ex-cattedra lectures on specific techniques (e.g., linear regression, linear discriminant analysis, clustering, etc.). Supervised and unsupervised learning are discussed in the framework of classification and clustering problems. The course outline is:

• Machine Learning and Pattern Classification: the general concepts of Machine Learning and Patter Recognition are introduced with a brief review of Statistical Decision Theory;
• Linear Regression Techniques: linear methods for regression will be disccussed and compared (e.g., Linear Regression and Ridge Regression).
• Linear Classification Techniques: linear methods for classification will be presented as the starting point for more complex methods (e.g., Linera Regression on the indicator matrix, Linear and Quadratic Discriminant Analysis, Logistic Regression, Percptron rule and Optimal Separating Hyperplanes, a.k.a., Support Vector Machines)
• Unsupervised Learning Techniques: the most common approaches to unsupervised learning are described mostly focusing on clustering techniques such as hierarchical clustering, k-means, k-medoids, Mixture of Gaussians, DBSCAN, Jarvis-Patrick, etc.;
• Model Validation and Selection: model validation and selection are orthogonal issues to all previous techniques; during the course their fundamentals are described and discussed in the framework of linear models for regression (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 Tuesday, in room V.S8-B, starts at 8:30 (quarto d'ora accademico), ends at 10:15
* On Fridays, in room V.S8-A, starts at 10:30 (quarto d'ora accademico), ends at 12:15 or 13:15 (check!)

Chapters are intended as complete except for

• Ch.4 ESL: Section 4.5
• Ch.12 ESL: Sections 12.1, 12.2, 12.3
• Ch.9 ISL: Sections 9.1, 9.2, 9.3

Course Evaluation

The 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

Homeworks

Please keep in mind:

• the homework is not meant to get more grade, it is intended for you to understand better, i.e., from a practical perspective too, the topics of the course
• the deadline to turn the homework in is the first time you take the PAMI exam, we will grade it when grading your classwork
• you can make the homework in groups up to 2/3 people, the deadline to turn it in is the date the first person in the groups takes the exam (and it will be graded for all members in the group at that call)
• provided you attended the lab lectures, and you have the R environment set up, the homework should require not more than 1 day per part

Homework 2016/2017

We have published the Homework 2016/2017.

Homework 2015/2016

We have published the Homework 2015/2016.

Homework 2014/2015

We have published the Homework 2014/2015.

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.

Lectures:

• [2016] Course introduction: introductory slides of the course with useful information about the grading, and the course logistics. Some examples from supervised and unsupervised learning. Regression, classification, clustering terminology and examples.
• [2016] Statistical Learning Introduction: Statistical Learning definition, rationale, and trade-offs (e.g., prediction vs. inference, parametric vs non parametric models, flexibility vs. interpretability, etc.)
• [2016] Statistical Learning and Model Assessment: Model Assessment for Regression and Classification, Bias-Variance trade-off, Model complexity and overfitting, K-Nearest Neighbors Classifier vs. Bayes Classifier.
• [2016] Linear Regression: Simple Linear Regression and Multiple Linear Regression. Feature selection. Ridge Regression and Lasso.
• [2016] Linear Classification: From Linear Regression to Logistic Regression. Linear Discriminant Analysis and Quadratic Discriminant Analysis. Comparison between linear classification methods.
• [2016] Support Vector Machines: Discriminative vs. generative methids. Hyperplanes learning and Perceptron. Maximum Margin Classifiers. The Kernel trick and Support Vector Machines.

For exercises and lab material please refer to Davide Eynard website.

Additional Resources

Papers and links useful to integrate the textbook

• Bias vs. Variance: "Understanding the Bias-Variance Tradeoff" essay by Scott Fortmann-Roe
• Karush Kuhn Tucker Conditions: a short note on their meaning with references to relevant wikipedia pages

Past Exams and Sample Questions

For some samples of exams you can check the last year PAMI ones

• 03/02/2016 PAMI Exam
• 19/02/2016 PAMI Exam
• 06/07/2016 PAMI Exam
• 09/09/2016 PAMI Exam
• 28/09/2016 PAMI 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)