Learning from Structured Data

Timetable: The course outline is shown below. Lectures are on Mondays 3-5pm in MVB 1.11a. There are no supervised labs (although lab space is available on Wednesdays 11am-1pm in MVB 2.11).

Textbook: The course is largely based on the textbook is Logical and Relational Learning by Luc De Raedt. The chapters corresponsing to each lecture are indicated below.

Week 13	Introduction to learning from structured data	Chap 1
Week 14	Foundations of logical learning and inference	Chap 2
Week 15	Representations for learning and mining	Chap 4	Assignment 1(20%) Deadline: 12th March
Week 16	Propositionalisation and flattening	Chap 4
Week 17	Generalisation and specialisation	Chap 5
Week 18	Inductive logic programming and Progol
Week 19	Upgrading propositional learners	Chap 6	Assignment2 (30%) Deadline: 30 April
Week 20	NO LECTURE (Coursework Preparation)
	Easter break
Week 21	Naive Bayesian classification of structured data
Week 22	Kernels and distances for structured data	Chap 9
Week 23	Conclusion
Week 24	NO LECTURE (Exam Revision)

Assessment:50% of the course mark comes from two assignments, which will be made accessible from the table above. The other 50% comes from the exam.

Materials: Printed materials will handed out at the start of each leacture and then posted on the table above. Supplementary materials for lectures not in the textbook will also be made available from the table above

Helpdesk: Helpdesk assistance is available from Tarek Abudawood on Wednesdays 11am-12am in MVB 1.09.

Lecturer: To discuss any other matters, please email Oliver Ray to arrange an appointment.

Feedback: The assignments on this unit are mostly open ended, so they will be marked by hand and there will be an overall marking report and a small amount of individual feedback for each one.