COMP24111 Course Unit Overview

Size: px

Start display at page:

Download "COMP24111 Course Unit Overview"

Beverly Bryant
5 years ago
Views:

1 COMP24111 Course Unit Overview Ke Chen and Tingting Mu

2 Introduction The Big Picture: Introductory machine learning course unit for 2 nd Year UG students Reasonable Math background required Matlab programming language used in lab exercises Contact time: 20-hour lectures and 10-hour lab sessions 10 two-hour lectures (11:00-13:00, Tuesday, Weeks 1-5 & Weeks 7-11) 5 two-hour lab sessions (Weeks 1, 3, 5, 8 and 10) Self-revision and back-log clearing lab marking in Week 12 No lecture but providing the self-revision materials A two-hour lab session added for completing lab ex. Marking (last chance!) 2

3 Lecture and Lab Part I (Weeks 1-5): Dr. Tingting Mu Five lectures Week 1: Machine learning basics, Nearest neighbour classifier Week 2-3: Linear classification and regression Week 4: Support vector machine Week 5: Neural network and deep learning Three lab sessions Week 1: Lab Ex. 1 (Matlab programming) and marking Week 3: Lab Ex. 2 help desk Week 5: Lab Ex. 2 marking 3

4 Lecture and Lab Part II (Weeks 7-12): Dr. Ke Chen Five lectures Week 7: Generative models and naïve Bayes Week 8: Clustering analysis basics Week 9: K-mean clustering Week 10: Hierarchical and ensemble clustering Week 11: Cluster validation Three lab sessions Week 8: Lab Ex. 3 help desk Week 10: Lab Ex. 3 marking Week 12: Clearing back-log (last chance for marking any of your Lab Ex.) 4

5 Examination (60%) Assessment Method Three sections: all questions are compulsory Section A: MCQs (30 marks); Q1-15 (Part I), Q16-30 (Part II) Section B: Questions pertaining to Part I (15 marks) Section C: Questions pertaining to Part II (15 marks) Lab Exercises (40%) Three lab exercises (Lab ex 2 & 3, the same deadline for all the groups) Exercise 1 (10 marks): Matlab programming (marked in your 1 st lab) Exercise 2 (15 marks): Face recognition (deadline: 11:00, 25th Oct. 2018) Exercise 3 (15 marks): Spam filtering (deadline: 11:00, 29th Nov. 2018) 5

6 Organise Your Time for Lab Work Exercise 1: Start as early as possible (start after this lecture). Exercise 2,3: Start as early as possible (start after the lecture in week 2 for ex2, start after the lecture in week 7 for ex3). Fully utilise the help desk session, be prepared in advance (week 3 for ex2, week 8 for ex3) Advice: To complete Ex 2 and Ex 3, you need to revise the knowledge learned in the lectures, and practise the knowledge through experiments. You have three weeks to work on each exercise. You should keep in mind that you will NOT be able to complete Ex 2 and 3 by working only in the allocated lab sessions. Start as early as possible! 6

7 Marking (Week 1,5,10,12): Lab Help and Marking Write your computer number in whiteboard when you need marking. All the lab ex marking takes place in Lab and will be marked by TAs under the supervision of lab supervisors (Part I: T. T. Mu, Part II: K. Chen). Remember to register attendance if not marked. Help Desk (Week 3, 8): Write your computer number in whiteboard when you have questions. If there is no question in the queue, TA can do marking. But please note that this is NOT prioritised in help desk. Remember to register attendance. 7

8 Other Information The teaching page (URL: syllabus.cs.manchester.ac.uk/ugt/comp24111/) contains all the information regarding this CU, e.g. lecture notes, lab ex. specification/deadline/policy, non-assessed ex, self-revision slides, FAQ, Read the FAQ page available on the teaching page (URL: before asking questions. Recommended textbooks [EA] E. Alpaydin, Introduction to Machine Learning (3 rd Ed.), MIT Press, (core) [KPM] K. P. Murphy, Machine learning: A Probabilistic Perspective, MIT Press, [CMB] C. M. Bishop, Pattern Recognition and Machine Learning, Springer,

Reducing Features to Improve Bug Prediction

Reducing Features to Improve Bug Prediction Shivkumar Shivaji, E. James Whitehead, Jr., Ram Akella University of California Santa Cruz {shiv,ejw,ram}@soe.ucsc.edu Sunghun Kim Hong Kong University of Science