ITEC203 Introduction to Data Science and Machine Learning
Unit rationale, description and aim
Artificial intelligence (AI) is the intelligence demonstrated by machines, devices, agents or computer programs, in addition to the natural intelligence displayed by humans and animals. AI is often considered as the study of intelligent and rational agents or machines that mimic cognitive functions associated with the human mind, such as problem solving, reasoning, planning, learning, actioning and decision making. Machine learning is a subfield of AI that studies the ability to improve machine performance based on experience. Machine learning (ML) employs algorithms and mathematical models that computer systems use to make decisions or predictions and it is prevalent in many contemporary AI applications that make common good and build better stewardship ranging from microelectronic devices to online services benefiting billions of users.
This unit will cover essential aspects of AI and ML, both theoretically and practically. This includes understanding, design and implementation of fundamental problem-solving algorithms in AI such as heuristic search and game theory as well as supervised and unsupervised ML algorithms. The aim of the unit is to learn essential concepts and techniques of AI and ML towards designing and building AI-enabled applications that makes people's lives better.
|Learning Outcome Number||Learning Outcome Description|
|LO1||Demonstrate comprehensive knowledge of ML programming with Python libraries and tools|
|LO2||Demonstrate both theoretical AI concepts and knowledge understanding and practical AI programming skills|
|LO3||Critically design an AI application that builds better stewardship with an appropriate choice of AI and ML techniques|
|LO4||Critically apply AI and ML techniques and tools to solve real-world problems that value human dignity|
Topics will include:
- Recap of ML basics and project design
- Bayesian statistics and information theory
- Support vector machines and k-nearest neighbors
- Decision trees, random forests and ensemble learning
- Unsupervised learning techniques
- Introduction to artificial neural networks
- Introduction to AI and intelligent agents
- Fundamental use cases for AI that makes common good and builds better stewardship
- Heuristic search techniques in AI
- Game theory in AI
- Overview of AI logic, knowledge, reasoning, planning and decision making
- AI ethics and safety with impact on the common good and human dignity
Learning and teaching strategy and rationale
This unit is offered in different modes to cater for the learning needs and preferences of a range of participants and maximise effective participation for isolated and/or marginalised groups. These are: “Attendance” mode, “Blended” mode and “Online” mode.
In a weekly attendance mode, students will require face-to-face attendance in specific physical location/s. Students will have face-to-face interactions with lecturer(s) or lab demonstrators to further their achievement of the learning outcomes. This unit is structured with required upfront preparation before workshops, most students report that they spend an average of one hour preparing before the workshop and one or more hours after the workshop practicing and revising what was covered. The online learning platforms used in this unit provide multiple forms of preparatory and practice opportunities for you to prepare and revise.
In a blended mode, students will require face-to-face attendance in blocks of time determined by the School and online asynchronous sessions for the other blocks of time during the semester. Students will have face-to-face interactions with lecturer(s) and online asynchronous engagement to further their achievement of the learning outcomes. This unit is structured with required upfront preparation before workshops. The online learning platforms used in this unit provide multiple forms of preparatory and practice opportunities for you to prepare and revise.
This unit uses an active learning approach to support students in the exploration of knowledge essential to the discipline. Students are provided with choice and variety in how they learn. Students are encouraged to contribute to asynchronous weekly discussions. Active learning opportunities provide students with opportunities to practice and apply their learning in situations similar to their future professions. Activities encourage students to bring their own examples to demonstrate understanding, application and engage constructively with their peers. Students receive regular and timely feedback on their learning, which includes information on their progress.
Students should anticipate undertaking 150 hours of study for this unit, including class attendance, readings, online engagement and forum participation and assessment preparation.
Assessment strategy and rationale
A range of assessment procedures will be used to meet the unit learning outcomes and develop graduate attributes consistent with University assessment requirements. The first assessment item is a programming practical that consists of several machine learning and AI programming tasks. The second assessment item is a Kaggle inclass competition to solve real-world problems via applying ML and AI techniques. It is used to assesses students’ AI and machine learning knowledge and understanding as well as practical AI programming skills with consideration of AI ethics and safety. The final assessment is Microsoft Azure AI Fundamentals certificate exam.
The assessments for this unit are designed to demonstrate the achievement of each learning outcome. To pass this unit, you are required to:
· obtain an overall mark of at least 50%
Overview of assessments
|Brief Description of Kind and Purpose of Assessment Tasks||Weighting||Learning Outcomes|
Assessment Task 1: programming practical
The first assessment item consists of several programming tasks. The purpose is to comprehensively assess students’ practical machine learning and AI skills using python and sophisticated machine learning/AI models. The deliverable is python code and solution/algorithm description.
Submission Type: Individual
Assessment Method: Practical Tasks
Artefact: Codes and comments in Jupyter Notebook
Assessment Task 2: A Kaggle Inclass competition
The second assessment item is to solve real-world problems via attempting ML/AI techniques in the format of Kaggle in-class competition. The deliverable is python code (with solution/algorithm description) and scores and rankings in the leaderborad of the competition.
Submission Type: Individual
Assessment Method: Kaggle inclass competition
Artefact: Codes and comments and scores and rankings in Kaggle inclass competition leaderboard.
Assessment Task 3: Microsoft Azure AI Fundamentals certificate exam.
The final assessment is a Microsoft certificate exam. The purpose is to assess students’ overall AI and machine learning knowledge and understanding in Microsoft certificate level.
Submission Type: Individual
Assessment Method: Exam
Representative texts and references
Aurélien Géron, 2019. Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition, O'Reilly Media, Inc.
Hadrien Jean, 2020. Essential Math for Data Science, O'Reilly Media, Inc.
Alberto Artasanchez and Prateek Joshi, 2020. Artificial Intelligence with Python, 2nd edition Packt Publishing.
Stuart Russell and Peter Norvig, 2020. Artificial Intelligence: A Modern Approach, 4th edition Pearson.
Richard E. Neapolitan and Xia Jiang, 2018. Artificial Intelligence: With an Introduction to Machine Learning, 2nd edition Chapman and Hall/CRC.