ST503 - Fundamentals of Linear Models and Regression
- Prerequisites: ST 501, ST 502 is a co-requisite
- Term & Frequency: Fall
- Student Audience: ST 503 is designed for students seeking a Masters of Statistics This course might also be taken by students from other disciplines who currently take ST 705, such as Financial Mathematics and Bioinformatics. The main pedagogical difference between the new ST 503 and ST 705 will be a de-emphasis on proofs and a new emphasis on real examples and data analysis to supplement the theory.
- Credit: 3 credits
- Recent Texts: Linear Models with R, by J. Faraway, 2nd edition (2014)
- Recent Instructors: Howard Bondell
- Background and Goals: The course covers the theory underlying linear statistical models as well as practical experience with model-building methods such as residual analysis and variable selection.
- Content: Estimation and testing in full and non-full rank linear models. Normal theory distributional properties. Least squares principle and the Gauss-Markoff theorem. Estimability, analysis of variance and covariance in a unified manner. Practical model building in linear regression including residual analysis, regression diagnostics, and variable selection. Emphasis on use of the computer to apply methods with data sets.
- Alternatives: ST 705. Credit not given for both ST705 and ST503. ST 708 (for students requiring only familiarity with aplied topics).
- Subsequent Courses: None
S1 2017 Sections:
|-||/ - |