Machine Learning & AI Foundations: Linear Regression

Expand your data science skills by learning how to leverage the concepts of linear regression to solve real-world problems.

Introduction

• Linear regression for machine learning
• What you should know
• Using the exercise files

1. Simple Linear Regression

• Building effective scatter plots in Chart Builder
• Adding labels and spikes to a scatter plot
• Create a 3D scatter plot
• Create a bubble chart
• Residuals and R2
• Calculating and interpreting regression coefficients

2. Introduction to Multiple Linear Regression

• Challenges and assumptions of multiple regression
• Checking assumptions visually
• Checking assumptions with Explore
• Checking assumptions: Durbin-Watson
• Checking assumptions: Levine's test
• Checking assumptions: Correlation matrix
• Checking assumptions: Residuals plot
• Checking assumptions: Summary

3. Dummy Code and Interaction Terms

• Creating dummy codes
• Dummy coding with the R extension
• Detecting variable interactions
• Creating and testing interaction terms

4. Three Regression Strategies

• Three regression strategies and when to use them
• Understanding partial correlations
• Understanding part correlations
• Visualizing part and partial correlations
• Simultaneous regression: Setting up the analysis
• Simultaneous regression: Interpreting the output
• Hierarchical regression: Setting up the analysis
• Hierarchical regression: Interpreting the output
• Creating a train-test partition in SPSS
• Stepwise regression: Setting up the analysis
• Stepwise regression: Interpreting the output

5. Spotting Problems and Taking Corrective Action

• Collinearity diagnostics
• Dealing with multicollinearity: Factor analysis/PCA
• Dealing with multicollinearity: Manually combine IVs
• Diagnosing outliers and influential points
• Dealing with outliers: Studentized deleted residuals
• Dealing with outliers: Should cases be removed?
• Detecting curvilinearity

6. Other Approaches to Regression

• Regression options
• Automatic linear modeling
• Regression trees
• Time series forecasting
• Categorical regression with optimal scaling
• Comparing regression to Neural Nets
• Logistic regression
• SEM

7. Advanced Alternatives Using the Extension Hub

• What is the extension hub?
• Ridge regression
• Lasso and elastic net

Conclusion

• What's next