Professor of Mathematics Lani Gao is integrating the use of artificial intelligence into her spring 2024 Math 5160 course, Applied Statistics Methods.
The primary objective of a new module she is adding to the course, “AI in Statistics Analysis,” is to integrate artificial intelligence by incorporating a module on turning cutting-edge technologies in AI into a statistical analysis tool.
The approach is introducing foundational concepts of AI relevant to statistical analysis such as machine learning, deep learning, neural networks and large language models such as ChatGPT.
AI-assisted data analysis is teaching students to leverage AI for data cleaning, imputation and examples of R codes. Discussions of machine learning and predictive modeling are intended to help students compare machine learning with traditional statistical methods.
The module is designed to enable students to demonstrate the ability to integrate AI into statistical analysis, improving efficiency. Students are expected to interpret the result of AI-enhanced statistical analysis in a clear and understandable manner, and to critically evaluate the strengths and limitations of AI-driven models compared to classical statistical methods in specific areas.
Students will complete a comprehensive project based on the learning module to assess their ability to integrate AI into statistical analysis.
For the project, multiple data sets will be provided, and each group of students will choose the data set of their preference. Students are also encouraged to use their own research data if possible.
Students are required to apply AI techniques for data preprocessing, including data cleaning, imputation and visualization. Students will compare the results with transitional preprocessing and discuss the impact on the data quality. Students are encouraged to seek examples of R code provided by AI tools which provide AI-enabled adaptive learning.
Students are required to implement both traditional statistical models and AI-driven models (such as machine learning algorithms) for predictive modeling.
Students will interpret the results of statistical analysis, emphasizing the differences between traditional and AI-driven approaches, discuss the implications of the findings, then discuss the interpretation ability of large language models such as ChatGPT.
College of Arts and Sciences Dean Pam Riggs-Gelasco funded Gao’s proposal to incorporate AI into a course module with $500 Gao has used to purchase online textbooks and course material and to attend a training workshop focused on AI and statistics to enhance her own expertise.