Courses Detail Information
STAT4060J – Computational Methods for Statistics and Data Science
Instructors:
Mark Fredrickson
Credits:
4
Pre-requisites:
MATH2160J/256/286，ECE4010J，ENGR1010J
Description:
This course introduces the art and practice of computational methods in statistics and data science. Students will learn how computational thinking underpins modern statistical analysis and data-driven discovery. Using R, Python, and C++, the course emphasizes hands-on implementation of core algorithms in statistical computing, including numerical optimization, simulation, and data modeling.
By combining algorithmic insight with practical coding experience, students will develop the skills to translate statistical concepts into efficient computational solutions applicable to real-world data problems.
Course Topics:
1. Matrix algebra
2. R Basics and R library
3. Least squares regression, sweep operator
4. Matrix decomposition
5. Logistic regression, Newton-Raphson
6. Adaboost, coordinate descent
7. Ridge regression, Spline regression
8. Lasso, stagewise regression
9. Feed-forward neural network, back-propagation
10. TensorFlow, Pytorch
11. Deep learning (Resnet, Transformer etc..)