3 - Probability: Multivariate Models
3 - Probability: Multivariate Models
3.1 - Joint distributions for multiple random variables
3.1.4 - Correlation does not imply causation
3.1.5 - Simpson’s paradox
3.2 - The multivariate Gaussian (normal) distribution
3.2.2 - Mahalanobis distance
3.2.3 - Marginals and conditionals of an MVN *
3.2.4 - Example: conditioning a 2d Gaussian
3.2.5 - Example: Imputing missing values *
3.3 - Linear Gaussian systems *
3.3.1 - Bayes rule for Gaussians
3.3.3 - Example: Inferring an unknown scalar
3.3.4 - Example: inferring an unknown vector
3.3.5 - Example: sensor fusion
3.4 - The exponential family *
3.4.3 - Log partition function is cumulant generating function
3.4.4 - Maximum entropy derivation of the exponential family
3.5 - Mixture models
3.5.1 - Gaussian mixture models
3.5.2 - Bernoulli mixture models
3.6 - Probabilistic graphical models *
