This course is best suited for individuals who have a technical background in mathematics/statistics/computer science/engineering pursuing a career change to jobs or industries that are data-driven such as finance, retain, tech, healthcare, government and many more. The opportunity is endless. In this course, we will learn what happens to our regression model when these assumptions have not been met. How can we detect these discrepancies in model assumptions and how do we remediate the problems will be addressed in this course.
Class Deals by MOOC List - Click here and see Coursera's Active Discounts, Deals, and Promo Codes.
This course is part of the Performance Based Admission courses for the Data Science program.
This course is the continuation of MAT764. If you have not yet taken the MAT764 course, it is recommended that you complete that course prior to this course. The foundational knowledge to support the project are carried through in this deeper dive into using core ideas behind simple and multiple linear regression assuming that all basic assumptions of the model have been met.
Upon successful completion of this course, you will be able to:
-describe the assumptions of the linear regression models.
-use diagnostic plots to detect violations of the assumptions of a linear regression model.
-perform a transformation of variables in building regression models.
-use suitable tools to detect and remove heteroscedastic errors.
-use suitable tools to remediate autocorrelation.
-use suitable tools to remediate collinear data.
-perform variable selections and model validations.
What you'll learn
- Describe the assumptions of the linear regression models.
- Use diagnostic plots to detect violations of the assumptions of a linear regression model.
- Perform variable selections and model validations.
Syllabus
Module 1: Model Diagnostics and Remediation Part I
Welcome to Model Diagnostics and Remediation Measures! In this course, we will cover the topics of: Regression Diagnostics, Variance Stabilizing Transformations, Box-Cox Transformation, Transformations to Linearized the Model, Weighted Least Squares, Autocorrelation, Multicollinearity, Variable Selection and Model Validation. In Module 1, we will cover four topics including: Regression Diagnostics, Variance Stabilizing Transformations, Box-Cox Transformation and Transformations to Linearize the model. There is a lot to read, watch, and consume in this module so, let’s get started!
Module 2: Model Diagnostics and Remediation Part II
Welcome to Module 2 – This module will cover four topics including: Weighted Least Squares, Autocorrelation, Multicollinearity, and Variable Selection and Model Validation. There is a lot to read, watch, and consume in this module so, let’s get started!
Summative Course Assessment
