Hello, if you have any need, please feel free to consult us, this is my wechat: wx91due
ACTL3143 Artificial Intelligence and Deep Learning Models for Actuarial Applications
Course Details & Outcomes
Course Description
Advances in artificial intelligence and machine learning - especially in deep learning methods - are creating products and services with the potential to change the insurance industry and actuarial work. This course will introduce students to concepts in Artificial Intelligence (AI), with particular focus on deep learning models, and their applications to risk and insurance. A particular focus of the course will be on how those AI models can be combined with other actuarial techniques in order to solve business problems in insurance and risk management including pricing, reserving and capital management, and insurance business processes. Students will be expected to understand the theory behind the models considered and their relationship to other actuarial techniques, and to fit and evaluate various deep learning models using appropriate software.
Course Aims
There are two main differences between this course and generic computer science treatment of deep learning. Firstly, the course will highlight deep learning models which can be combined with other actuarial techniques (e.g. claim frequency modelling with GLMs, mortality forecasting). Secondly, the assumed knowledge for the course is tailored to actuarial students. Only a minimal coding experience is assumed (e.g. the basics of variables, control flow, functions), and the required coding concepts in Python will be taught in the lectures.
This course will complement the machine learning methods covered in ACTL3142 Actuarial Data and Analysis.
Course Learning Outcomes
| Course Learning Outcomes | Program learning outcomes |
|---|---|
| CLO1 : Develop an understanding of artificial intelligence techniques and their potential application in insurance business processes |
|
| CLO2 : Understand and explain the key features of various deep neural networks and their applications to various specialist areas of actuarial studies and insurance tasks, including highlighting their differences and limitations |
|
| CLO3 : Implement deep neural networks using deep learning software and real world datasets |
|
| CLO4 : Combine deep neural networks with other actuarial techniques to solve problems in various types of insurance business, including pricing, reserving, and capital management |
|
| CLO5 : Apply effective communication, discussion and report writing skills when interpreting results from deep learning algorithms and artificial intelligence systems |
|
| CLO6 : Understand the ethical considerations related to the data colllection, design and implementation of various AI and deep learning models for actuarial and insurance applications |
|
| Course Learning Outcomes | Assessment Item |
|---|---|
| CLO1 : Develop an understanding of artificial intelligence techniques and their potential application in insurance business processes |
|
| CLO2 : Understand and explain the key features of various deep neural networks and their applications to various specialist areas of actuarial studies and insurance tasks, including highlighting their differences and limitations |
|
| CLO3 : Implement deep neural networks using deep learning software and real world datasets |
|
| CLO4 : Combine deep neural networks with other actuarial techniques to solve problems in various types of insurance business, including pricing, reserving, and capital management |
|
| CLO5 : Apply effective communication, discussion and report writing skills when interpreting results from deep learning algorithms and artificial intelligence systems |
|
| CLO6 : Understand the ethical considerations related to the data colllection, design and implementation of various AI and deep learning models for actuarial and insurance applications |
|
Learning and Teaching Technologies
Moodle - Learning Management System | Zoom | Echo 360 | EdStem
Learning and Teaching in this course
This course consists of:
- Self-study course material available on the course Moodle website (e.g. textbook chapters, lecture notes, exercises),
- Weekly lectures,
- Weekly labs, and
- Weekly consultation times.
Assessments
Assessment Structure
| Assessment Item | Weight | Relevant Dates | Program learning outcomes |
|---|---|---|---|
|
Storywall Discussion forums
Assessment FormatIndividual
|
30% |
Start DateWeek 1
Due DateThroughout term, see Moodle for details
|
|
|
Assignment
Assessment FormatIndividual
|
40% |
Start DateWeek 2
Due DateThe milestones will be due around weeks 5, 8, and 9 respectively, though see the Moodle page for the specific dates.
|
|
|
Final Exam
Assessment FormatIndividual
|
30% |
Start DateNot Applicable
Due DateNot Applicable
|
|
Assessment Details
-
Storywall Discussion forums
Assessment Overview
These are aimed at encouraging students to keep up with the course materials.
Course Learning Outcomes
- CLO1 : Develop an understanding of artificial intelligence techniques and their potential application in insurance business processes
- CLO2 : Understand and explain the key features of various deep neural networks and their applications to various specialist areas of actuarial studies and insurance tasks, including highlighting their differences and limitations
- CLO3 : Implement deep neural networks using deep learning software and real world datasets
- CLO4 : Combine deep neural networks with other actuarial techniques to solve problems in various types of insurance business, including pricing, reserving, and capital management
- CLO5 : Apply effective communication, discussion and report writing skills when interpreting results from deep learning algorithms and artificial intelligence systems
- CLO6 : Understand the ethical considerations related to the data colllection, design and implementation of various AI and deep learning models for actuarial and insurance applications
Detailed Assessment Description
The course offers formative activities to practice the concepts you have learned each week and aim at encouraging students to keep up with the course materials. These activities will reinforce your learning and help you identify the areas you need to focus on.
Submission notes
On Moodle
Assignment submission Turnitin type
Not Applicable
-
Assignment
Assessment Overview
Project involving application of course concepts.
Course Learning Outcomes
- CLO1 : Develop an understanding of artificial intelligence techniques and their potential application in insurance business processes
- CLO2 : Understand and explain the key features of various deep neural networks and their applications to various specialist areas of actuarial studies and insurance tasks, including highlighting their differences and limitations
- CLO3 : Implement deep neural networks using deep learning software and real world datasets
- CLO4 : Combine deep neural networks with other actuarial techniques to solve problems in various types of insurance business, including pricing, reserving, and capital management
- CLO5 : Apply effective communication, discussion and report writing skills when interpreting results from deep learning algorithms and artificial intelligence systems
- CLO6 : Understand the ethical considerations related to the data colllection, design and implementation of various AI and deep learning models for actuarial and insurance applications
Detailed Assessment Description
There will be a major assignment task involving the practical application of deep learning concepts in the course. It will also assess critical analysis and problem solving skills as well as communication skills, and correspond to course learning outcomes, and program learning goals.
The project will be submitted in stages:
1. Report Part 1 (10%),
2. Recorded Presentation (15%),
3. Report Part 2 (15%).
Assessment Length
5 pages
Submission notes
See lectures and Moodle for specific instructions.
Assignment submission Turnitin type
This assignment is submitted through Turnitin and students can see Turnitin similarity reports.
-
Final Exam
Assessment Overview
The examination will aim to assess the achievement of the learning course outcomes.
Course Learning Outcomes
- CLO1 : Develop an understanding of artificial intelligence techniques and their potential application in insurance business processes
- CLO2 : Understand and explain the key features of various deep neural networks and their applications to various specialist areas of actuarial studies and insurance tasks, including highlighting their differences and limitations
- CLO4 : Combine deep neural networks with other actuarial techniques to solve problems in various types of insurance business, including pricing, reserving, and capital management
- CLO5 : Apply effective communication, discussion and report writing skills when interpreting results from deep learning algorithms and artificial intelligence systems
- CLO6 : Understand the ethical considerations related to the data colllection, design and implementation of various AI and deep learning models for actuarial and insurance applications
Detailed Assessment Description
The exam will test the concepts presented in the lectures.
Submission notes
See lectures & Moodle for further details.
Assignment submission Turnitin type
Not Applicable
General Assessment Information
As a student at UNSW you are expected to display academic integrity in your work and interactions. Where a student breaches the UNSW Student Code with respect to academic integrity, the University may take disciplinary action under the Student Misconduct Procedure. To assure academic integrity, you may be required to demonstrate reasoning, research and the process of constructing work submitted for assessment.
To assist you in understanding what academic integrity means, and how to ensure that you do comply with the UNSW Student Code, it is strongly recommended that you complete the Working with Academic Integrity module before submitting your first assessment task. It is a free, online self-paced Moodle module that should take about one hour to complete.
Grading Basis
Standard
Requirements to pass course
In order to pass this course students must:
- Achieve a composite mark of at least 50 out of 100
- Engage actively in course learning activities and attempt all assessment requirements
- Meet any additional requirements specified in the assessment details
- Meet the specified attendance requirements of the course
Course Schedule
| Teaching Week/Module | Activity Type | Content |
|---|---|---|
| Week 1 : 27 May - 2 June | Lecture |
Artificial Intelligence, Neural Networks, and Python |
| Week 2 : 3 June - 9 June | Lecture |
Keras Deep Learning for Tabular Data |
| Week 3 : 10 June - 16 June | Lecture |
Computer Vision |
| Week 4 : 17 June - 23 June | Lecture |
Natural Language Processing |
| Week 5 : 24 June - 30 June | Lecture |
Recurrent Neural Networks |
| Week 6 : 1 July - 7 July | Other |
Flexibility week - no classes |
| Week 7 : 8 July - 14 July | Lecture |
Advanced Topics |
| Week 8 : 15 July - 21 July | Lecture |
Advanced Topics |
| Week 9 : 22 July - 28 July | Lecture |
Advanced Topics |
| Week 10 : 29 July - 4 August | Lecture |
Advanced Topics and Exam Preparation |
Attendance Requirements
Students are strongly encouraged to attend all classes and review lecture recordings.
Course Resources
Prescribed Resources
Course website
The website for this course is on Moodle.
The course will use various digital resources, but they all will be linked from Moodle.
To access the Moodle online support site for students, follow the links from that website to UNSW Moodle Support/Support for Students. Additional technical support can be obtained from [email protected] (02 9385 1333).
All course contents will be available from the course website. It is essential that you visit the site regularly to see any notices posted there by the course coordinator, as it will be assumed that they are known to you within a reasonable time.
Textbooks
There are many books of relevance to the course topics. The following book will be the main text references for a substantial part of the course: Aurélien Géron's textbook Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow (2nd or 3rd Edition), available digitally via UNSW Library's O'Reilly subscription.
Additional readings from the professional actuarial literature will also be used to provided additional context, details, and examples. This will be communicated in the course website.
Course Evaluation and Development
Feedback is regularly sought from students and continual improvements are made based on this feedback. At the end of this course, you will be asked to complete the myExperience survey, which provides a key source of student evaluative feedback. Your input into this quality enhancement process is extremely valuable in assisting us to meet the needs of our students and provide an effective and enriching learning experience. The results of all surveys are carefully considered and do lead to action towards enhancing educational quality.