31005 Machine Learning Subject Outline PDF

Preview

Summary

This is the subject outline for 31005 Machine Learning and 32513 Advanced Data Analytics Algorithms...

Description

SUBJECT OUTLINE 31005 Machine Learning Course area

UTS: Information Technology

Delivery

Spring 2021; City

Credit points 6cp Requisite(s)

31250 Introduction to Data Analytics AND 48024 Applications Programming

Result type

Grade and marks

Attendance: Forms of attendance and mode of delivery in this subject have changed to enable social distancing and reduce the risks of spreading COVID-19 in our community. Recommended studies: knowledge of database technologies

Subject coordinator Name: Dr. Jun Li E-mail: [email protected]

All email sent to subject coordinators, tutors or lecturers must have a clear subject line that states the subject numbe followed by the subject of the email [e.g. Subject 31005, Request for Extension], and must be sent from your UTS email address.

Teaching staff The subject will contain a mix of teaching by the following lecturer and guest lectures by others. Dr. Jun Li - email: [email protected]

Subject description

Machine learning is an exciting field studying of how intelligent agents can learn from and adapt to experience and how to realise such capacity on digital computers. It is applied in many fields of business, industry and science to discover new information and knowledge. At the heart of machine learning are the knowledge discovery algorithms. This subject builds on previous data analytics subjects to give an understanding of how both basic and more powerfu algorithms work. It consists of both hands-on practice and fundamental theories. Students learn important techniques in the field by implementation and theoretical analysis. The subject also introduces practical applications of machine learning, especially in the field of artificial intelligence.

Subject learning objectives (SLOs) Upon successful completion of this subject students should be able to: 1. describe the scope, limitations and application of several advanced data machine learning methods; 2. use or program a machine learning method; 3. design an approach to machine learning problems in specialised domains; 4. demonstrate an understanding of the issues underlying machine learning to successfully outline an approach to solving a machine learning problem.

Course intended learning outcomes (CILOs)

This subject also contributes specifically to the development of the following Course Intended Learning Outcomes (CILOs): Socially Responsible: FEIT graduates identify, engage, interpret and analyse stakeholder needs and cultural perspectives, establish priorities and goals, and identify constraints, uncertainties and risks (social, ethical, cultur legislative, environmental, economics etc.) to define the system requirements. (B.1) Design Oriented: FEIT graduates apply problem solving, design and decision-making methodologies to develop components, systems and processes to meet specified requirements. (C.1) 27/07/2021 (Spring 2021)

© University of Technology Sydney

Page

Collaborative and Communicative: FEIT graduates work as an effective member or leader of diverse teams, communicating effectively and operating within cross-disciplinary and cross-cultural contexts in the workplace. (E.1)

Teaching and learning strategies

The subject is delivered by online learning materials (organised videos of short lectures and algorithm implementatio tutorials) and interactive workshops, as well as industry-based guest lectures. The subject features in-depth study of the theory and algorithm of data analytics, as well as detailed hands-on implementation tutorials of classical algorithms. Guest lectures also highlight UTS-specific and industry-based research that give students the opportunity to engage deeply with experts and ask questions that address advanced methods in data analytics. Students will engage with pre-reading material that will be used as basis for discussion and activities in class. Each week an in-cla test is made available for students to check their knowledge and gauge their strengths and areas needing further practice. In-class tests with immediate feedback will help students to reflect on their learning. In this subject students have the opportunity to prepare a firm foundation for further study in data science and artificial intelligence by engagi deeply with the project.

Content (topics) 1. 2. 3. 4. 5.

Machine learning and relationship to statistics and artificial intelligence Theory of learning from data Important learning models Information theory Evaluation method and decision making

Program Week/Session

Dates

Description

1

2 Aug

Introduction to the subject Machine Learning Problem

2

9 Aug

Machine Learning Framework and 4 Elements

3

16 Aug

Linear Family of Data Models

4

23 Aug

Learning Algorithm -- A Case Study of Perceptron Training

5

30 Aug

Decision Trees Motivated from an Information Theoretic Viewpoint

6

6 Sep

Decision Tree Building Algorithm

7

13 Sep

Model Evaluation

StuVac

20 Sep

StuVac

8

27 Sep

Bootstrapping and Ensemble Methods

9

4 Oct

Neural Networks 1: Definition and Forward Computation

27/07/2021 (Spring 2021)

© University of Technology Sydney

Page

10

11 Oct

Neural Networks 2: Backward propagation algorithm

11

18 Oct

Guest lecture Notes: Topic will be announced later during the class

12

25 Oct

Guest lecture Notes: Topic will be announced later during the class

Assessment Assessment task 1: Quizzes Objective(s): This assessment task addresses the following subject learning objectives (SLOs): 1 and 4 This assessment task contributes to the development of the following Course Intended Learning Outcomes (CILOs): B.1 and C.1 Type:

Quiz/test

Groupwork:

Individual

Weight:

30%

Task:

Quizzes are presented as multiple choice, 2-4 questions per week for ten weeks. These are completed in class and students receive immediate feedback.

Due:

In-class

Assessment task 2: Algorithm Implementation and Report Objective(s): This assessment task addresses the following subject learning objectives (SLOs): 1, 2, 3 and 4 This assessment task contributes to the development of the following Course Intended Learning Outcomes (CILOs): C.1 and E.1 Type:

Project

Groupwork: Individual Weight:

50%

27/07/2021 (Spring 2021)

© University of Technology Sydney

Page

Task:

The goal of this assignment is to develop your hands-on skills in performing learning from data, as well as further understanding of the practical technical details of the learning procedure. You will implement a simple machine learning algorithm from scratch, for example, the ID3 decision tree building algorithm or the perceptron training algorithm or an ensemble method. You can also implement another algorithm of your interest to solve the supervised learning problem, with similar levels of details as demonstrated in tutorial videos (approx 2-3 hour tutorial video demoed by the instructor.) The algorithm needs to be tested using at least one simple but practical dataset and under an appropriate learning framework as instructed in the course. Then write a report about the implementation: model introduction, interface, algorithm and data structure design, testing/evaluation strategy and results. Alternatively, unsupervised learning algorithms can also be considered, but you must specify the testing scheme and the criteria clearly in the report (see below) if you choose to implement an unsupervised learning algorithm. The details about submission/evaluation/minimal datasets will be released in the assignment specification documentation.

Length:

A report of about 1000 words

Due:

11.59pm Sunday 3 October 2021

Assessment task 3: Video Presentation Objective(s): This assessment task addresses the following subject learning objectives (SLOs): 1 and 2 This assessment task contributes to the development of the following Course Intended Learning Outcomes (CILOs): C.1 and E.1 Type:

Presentation

Groupwork:

Individual

Weight:

20%

Task:

You will be randomly given 3 quiz questions in Assessment Task 1. Then you need to submit a short video presentation (max 5:00) of your explanation or reflection about your choices of each of the questions (and why you had made errors in Assessment Task 1, if applicable).

Due:

11.59pm Sunday 17 October 2021

Moderation of marks Where assessment items are marked by more than one marker, moderation will occur in line with UTS policy.

Minimum requirements In order to pass the subject, a student must achieve an overall mark of 50% or more.

Recommended texts You might find the following texts useful. 1. Discovering Knowledge in Data, D. T. Larose and C. D. Larose, Wiley, 2014. 2. Learning from Data, Y. S. Abu-Mostafa, M. Magdon-Ismail and H-T. Lin, AMLbook.com, 2016. 3. Introduction to Data Mining, P.-N. Tan, M. Steinbach and V. Kumar, Addison-Wesley, 2005. 27/07/2021 (Spring 2021)

© University of Technology Sydney

4. Machine Learning, Tom M Mitchell, McGraw-Hill, 1997. 5. Pattern Recognition and Machine Learning, Christopher Bishop, Springer, 2006. 6. Data Mining: Concepts and Techniques, J. Han and M. Kamber, Morgan Kaufmann, 2001.

References The UTS Coursework Assessment Policy & Procedure Manual, at www.gsu.uts.edu.au/policies/assessment-coursework.html.

Page...