Undergraduate Course: Machine Learning and Pattern Recognition (INFR11130)
Course Outline
School  School of Informatics 
College  College of Science and Engineering 
Credit level (Normal year taken)  SCQF Level 11 (Year 4 Undergraduate) 
Availability  Available to all students 
SCQF Credits  20 
ECTS Credits  10 
Summary  Machine learning is a field devoted to developing algorithms that adapt their behaviour to data, providing useful representations of the data and/or predictions. This course covers some fundamental theoretical concepts in machine learning, and common patterns for implementing methods in practice. The intended audience are those wanting the background required to begin research and development of machine learning methods.
This is an advanced course. Students should not choose this class without the required background (see "Other Requirements" box); students without this background are strongly advised to take a more practicalbased course, such as Introduction to Applied Machine Learning (INFR10063).
This 20 credit course replaces Machine Learning & Pattern Recognition (Level 11) (INFR11073)  10 credit course. 
Course description 
The precise set of methods and algorithms used to illustrate and explore the main concepts will change slightly from year to year. However, the main topic headings are expected to be fairly stable.
 Classification and Regression:
Linear Regression, logistic regression, Bayes classifiers
 Expanded feature representations:
Basis functions, neural networks, kernel methods
 Generalization, regularization and inference:
Penalized cost functions, Bayesian prediction, learning theory
 Model selection, pruning and combination:
Crossvalidation, Bayesian methods, sparsifying regularizers, ensemble methods.
 Representation and metric learning:
dimensionality reduction, clustering, feature learning
To support these topics we will also cover:
 Optimization and Inference algorithms:
Stochastic gradient descent, simple Monte Carlo ideas, and more specialized methods as required.
Practical issues:
 Formulating problems as machine learning, adapting methods to fit problems.
 Numerical and programming issues important for machine learning.
 Ethical issues, such as responsible application of methods and privacy concerns.

Entry Requirements (not applicable to Visiting Students)
Prerequisites 

Corequisites  
Prohibited Combinations  Students MUST NOT also be taking
Machine Learning & Pattern Recognition (Level 11) (INFR11073)

Other requirements  This course is open to all Informatics students including those on joint degrees. For external students where this course is not listed in your DPT, please seek special permission from the course organiser.
This course requires practical mathematical application of algebra, vectors and matrices, calculus, probability, and problem solving. For example, you will need to be able to differentiate linear algebra expressions with respect to vectors, interpret innerproducts and quadratic forms geometrically, and compute expectations of linear algebra expressions under simple distributions. Some of the required details can be learned during the course. However, practical mathematical skills take time to accumulate and a strong mathematical background is essential.
Practical exercises usually require using a particular numerical language such as Matlab or Python+NumPy. We will assume and require sufficient past programming experience that a new package can be learned on the fly. 
Information for Visiting Students
Prerequisites  See above. 
High Demand Course? 
Yes 
Course Delivery Information

Academic year 2016/17, Available to all students (SV1)

Quota: None 
Course Start 
Semester 1 
Timetable 
Timetable 
Learning and Teaching activities (Further Info) 
Total Hours:
200
(
Lecture Hours 30,
Seminar/Tutorial Hours 10,
Feedback/Feedforward Hours 2,
Summative Assessment Hours 20,
Programme Level Learning and Teaching Hours 4,
Directed Learning and Independent Learning Hours
134 )

Assessment (Further Info) 
Written Exam
80 %,
Coursework
20 %,
Practical Exam
0 %

Additional Information (Assessment) 
Coursework 20%, Exam 80% 
Feedback 
Some of the lecture time will be devoted to discussing questions,
including some examlike questions, and providing feedback on student
answers. Students will also get feedback on their work through the
tutorials. 
Exam Information 
Exam Diet 
Paper Name 
Hours & Minutes 

Main Exam Diet S2 (April/May)   2:00  

Academic year 2016/17, Partyear visiting students only (VV1)

Quota: None 
Course Start 
Semester 1 
Timetable 
Timetable 
Learning and Teaching activities (Further Info) 
Total Hours:
200
(
Lecture Hours 30,
Seminar/Tutorial Hours 10,
Feedback/Feedforward Hours 2,
Summative Assessment Hours 20,
Programme Level Learning and Teaching Hours 4,
Directed Learning and Independent Learning Hours
134 )

Assessment (Further Info) 
Written Exam
80 %,
Coursework
20 %,
Practical Exam
0 %

Additional Information (Assessment) 
Coursework 20%, Exam 80% 
Feedback 
Some of the lecture time will be devoted to discussing questions,
including some examlike questions, and providing feedback on student
answers. Students will also get feedback on their work through the
tutorials. 
Exam Information 
Exam Diet 
Paper Name 
Hours & Minutes 

Main Exam Diet S1 (December)   2:00  
Learning Outcomes
On completion of this course, the student will be able to:
 Frame an applied problem as a machine learning task, identifying appropriate methods.
 Critically compare and contrast alternative machine learning methods for a given task.
 Derive and motivate novel variants of machine learning methods.
 Create accessible and useful explanations of the workings and failure modes of machine learning methods.
 Check and refine implementations of learning algorithms, while applying them in practice.

Reading List
Machine Learning: A Probabilistic Perspective. Kevin P Murphy.
 Bayesian Reasoning and Machine Learning. David Barber.
The following may also be of interest:
 Pattern Recognition and Machine Learning, Christopher Bishop.
 A First Course in Machine Learning (2nd Ed.), Simon Rogers and Mark
Girolami.
 Information Theory, Inference and Learning Algorithms, David MacKay
 The Elements of Statistical Learning: Data Mining, Inference, and
Prediction, Hastie, Tibshirani, and Friedman. 
Additional Information
Graduate Attributes and Skills 
The student will be able to reason about how to make predictions from and interpret data, an important transferable skill.
In addition the student will be able to:
Undertake critical evaluations of a wide range of numerical and graphical data.
Apply critical analysis, evaluation and synthesis to forefront issues, or issues that are informed by forefront developments in the subject/discipline/sector.
Identify, conceptualise, and define new and abstract problems and issues.
Develop original and creative responses to problems and issues.
Critically review, consolidate and extend knowledge, skills, practices and thinking in subject/discipline/sector.
Deal with complex issues and make informed judgements in situations in the absence of complete or consistent data/information.

Keywords  Machine Learning,supervised learning,Probabilistic prediction,Data science 
Contacts
Course organiser  Dr Iain Murray
Tel: (0131 6)51 9078
Email: I.Murray@ed.ac.uk 
Course secretary  Mr Gregor Hall
Tel: (0131 6)50 5194
Email: gregor.hall@ed.ac.uk 

