Module: DBA3803 Predictive Analytics in Business
Semester taken: SEM1 AY2020/2021
Lecturer / Tutor: Long Zhao
Module Synopsis (for the semester that I took)
This course aims to develop an understanding of forecasting methods from data science for analyzing complex issues and solving business problems. We will make productive use of analytics tools available in R and Python. Because these packages are mature and convenient, we will instead focus on the thinking behind the methodology, which also applies to more advanced tools. Moreover, we will also learn the limitations of forecasting methods and common illusions in predictive analytics. Although the class focuses on simplified models, it aims to bridge the classroom knowledge and business applications, such as portfolio construction, e-commerce, and smart city operations.
Main Learning Objectives
- Understand data science principles
- Understand techniques and methods for predictive analytics
Course Deliverables / Graded Components
- 4 Group Projects: 50%
- Individual Assignments: 20%
- 2 Quizzes: 20%
- Attendance: 10%
I took this module as it was a compulsory module for the Business Analytics (BA) specialisation.
The module covered methods and techniques for predictive analytics, which included linear regression, regularization, overfitting, decision trees and classification. While there was a large proportion of hands-on practise through the DataCamp courses, the course focused on the science and methodology behind said methods. As such, the course did not shy away from equations, graphs and mathematical calculations.
The group projects were similar to case studies as the projects’ premise and discussion questions were given. The group projects and individual assignments were more application-based (i.e. use of R and/or Python), while the quizzes were more theory-based (i.e. testing of concepts), which I think was a fairly good mixture.
Regarding lesson format, class was conducted in lecture style. During lesson time, Prof Long Zhao explained about the different techniques. Outside of lesson time, students were expected to hone their technical skills via DataCamp courses on a weekly basis.
Personally speaking, I found it hard to grasp the methodology behind forecasting techniques and often zoned out during lessons. I believe that Prof Long Zhao has a solid understanding of the content, but, in my opinion, falls short in effective teaching. I had tried to self-learn the concepts via online research, but eventually I only managed to scrap through and obtain an average grade for the module.
This module is not for the faint-hearted. A solid foundation in math and analytical skills would be essential. Also, please find out which professor is teaching in the relevant semester as the course curriculum would differ.
Read my full list of module reviews!
I also sell my module notes here, feel free to browse!
Thank you for reading!