Data Science Project (390-0-20)
Instructors
Arvind Krishna
Meeting Info
Parkes Hall 215: Mon, Wed 2:00PM - 3:20PM
Overview of class
In this course, students will collaborate on real-world data science project introduced by a real stakeholder. Specifically, for the Spring 2025 Quarter, they will continue the work on the STAT390 project started in Fall 2024, which is to develop a classification model to classify the severity of potential eye cancer in a patient, based on medical image data. They will have the opportunity to meet the stakeholders - (1) Dr. Yamini Krishna, Consultant Ophthalmic Pathologist at the Royal Liverpool University Hospital, UK & Honorary Senior Clinical Lecturer, Eye & Vision Science Department, University of Liverpool, UK, and (2) Dr. He Zhao, Lecturer, Eye & Vision Science Department, University of Liverpool, UK. The stakeholder meetings will be scheduled based on the availability of the stakeholders, and the requirements of the project.
Students will work in teams, such as (A) Coding team, (B) Communications team, (C) Literature survey team, (D) Strategy team, and so on. The core part of the project will involve processing image data, feature engineering, developing and testing a deep learning classification model.
Using tools like Git and GitHub, students will develop teamwork skills while gaining hands-on experience on application of advanced machine learning methods. The course provides practical skills and exposure to modern data science tools and collaborative environments.
Registration Requirements
STAT 301-3 or STAT 303-3 or consent of instructor. Will be good to have STAT362 completed as well.
Learning Objectives
1. Experience the challenges and the open-ended nature of a real-data science project
2. Experience the opportunity to talk to the stakeholders of a real-data science project
3. Convert the problem between its domain-specific language and data science terminology
4. Learn the commonly used tools and techniques useful in a data science project
5. Perform exploratory data analysis (EDA) and apply modern machine learning methods
6. Develop proficiency in collaborative teamwork using version control systems like Git and GitHub
7. Demonstrate the ability to effectively document and present analysis results in a clear and organized manner.
Teaching Method
The students will assigned to teams based on their skills and preferences. Teams will present their progress on the weekly objectives once every week, receive feedback from the instructor, and objectives for next week. The weekly objectives will be decided by the instructor, but will be intended to achieve the stakeholders' goals.
Evaluation Method
A student in this class will be evaluated based on their contribution in achieving the weekly objectives of their team.
Class Materials (Required)
No required textbook.
Class Materials (Suggested)
In-class lecture notes will be provided
Enrollment Requirements
Enrollment Requirements: Prerequisites: STAT 301-3 or STAT 303-3