Duration
2 days
What you'll learn
- Understand what regression is
- Learn about different types of Regression Models and their use
- Build linear regression models
- Build logistic regression models
- Build count models
- Interpret regression results
- Visualise the results
- Test model assumptions
Course Outline
Introduction to Simple Statistics
- Sample And Population
- Central Tendency Measurements
- Normal Distribution Simple Linear Regression
Simple linear regression
- The slope
- R-squared
- The p-value
- Model fit
- The residuals
Multiple linear regression
- The slopes
- R-squared
- The p-value
- Model fit and residuals
Linear Regression: Binary, Categorical, and Quadratic Variables
- Binary variables
- Categorical variables
- Quadratic variables
Linear Regression: Checking Model Fit and Assumptions
- Prediction
- Normality of residuals
- Independence of residuals
- Constant variance
- Multi collinearity
- Outliers
- Influential observations
- Selection algorithms
Linear Regression Case Study
- The dataset
- Including continuous variables
- Including binary variables
- Including categorical variables
- Multiple regression
- Checking model fit
- Checking model assumptions
- Multicollinearity
- Outliers
- Influential observations
- Visualizing the result
Logistic Regression: Contingency Tables
- Two-by-two tables
- The odds
- The odds ratio
- Two-by-three tables
Logistic Regression Models
- Single independent variable
- Binary variables
- Multiple independent variables
- Categorical variables
- Nonlinearity: Non-graphical test
- Nonlinearity: Graphical test
Logistic Regression: Prediction and Model Fit
- Prediction
- Goodness of fit: Likelihood ratio test
- Goodness of fit: Hosmer-Lemeshow test
- Goodness of fit: Classification tables
- Goodness of fit: ROC analysis
- Residuals
- Influential Observations
Logistic Regression Case Study
- The dataset
- Continuous variables
- Test of linearity: Non-graphical
- Test of linearity: Graphical
- Binary variables
- Categorical variables
- Multivariate analysis
- Goodness of fit
- Residual analysis
- Influential observations
- Combining both residuals and influence in one graph
- Visualizing the result
Count Models: Count Tables
- Count tables
- Risk
- Incidence-rate ratio
- Two-by-three tables
Poisson Regression
- Single independent variable
- Binary variables
- Multiple independent variables
- Categorical variables
Other Count Models
- Negative binomial regression
- Truncated models
- Zero-inflated models
- Comparison of models
- Prediction
- Predicting the number of events
- Predicting probabilities of certain counts
Count Model Case Study
- The dataset
- Continuous variables
- Binary variables
- Multivariate analysis
- Negative binomial regression
- Zero-inflated models
- Comparing count models
- Visualizing the result
Programme scope, delivery arrangements and suitability are confirmed during enquiry.