## STATS 305A: Applied Statistics I

Statistics of real valued responses. Review of multivariate normal distribution theory. Univariate regression. Multiple regression. Constructing features from predictors. Geometry and algebra of least squares: subspaces, projections, normal equations, orthogonality, rank deficiency, Gauss-Markov. Gram-Schmidt, the QR decomposition and the SVD. Interpreting coefficients. Collinearity. Dependence and heteroscedasticity. Fits and the hat matrix. Model diagnostics. Model selection, Cp/AIC and crossvalidation, stepwise, lasso. Multiple comparisons. ANOVA, fixed and random effects. Use of bootstrap and permutations. Emphasis on problem sets involving substantive computations with data sets. Prerequisites: consent of instructor, 116, 200, applied statistics course,
CS 106A,
MATH 114. (NB: prior to 2016-17 the 305ABC series was numbered as 305, 306A and 306B).

Terms: Aut
| Units: 3

Instructors:
Owen, A. (PI)

## STATS 305B: Applied Statistics II

Methods for discrete responses. Logistic regression. Poisson loglinear models. Probit model. Generalized linear models. Exponential families. Overdispersion. Tests of independence. Cox proportional hazards model. Bradley-Terry models. Cochran-Mantel-Haenszel. Matched pairs. Generalized linear mixed models. Problem sets involve working with real data sets. Additional topics selected by the instructor. Prerequisites: 305A or consent of the instructor. (NB: prior to 2016-17 the 305ABC series was numbered as 305, 306A and 306B).

Terms: Win
| Units: 3

Instructors:
Taylor, J. (PI)

## STATS 305C: Applied Statistics III

Methods for multivariate responses. Theory, computation and practice for multivariate statistical tools. Multivariate Gaussian and undirected graphical models, graphical displays. Hotelling¿s T-squared, principal components, canonical correlations, linear discriminant analysis, correspondence analysis, and recent variants of these. Hierarchical and k-means clustering. Bi-clustering. Factor analysis and independent component analysis. Topic modeling. Multidimensional scaling and variants (e.g., Isomap, spectral clustering, t-SNE). Matrix completion. Extensive work with data involving programming ¿ ideally in R. nPrerequisites:
Stats 305A and
Stats 305B or consent of the instructor. (NB: prior to 2016-17 the 305ABC series was numbered as 305, 306A and 306B).

Terms: Spr
| Units: 3

Instructors:
Holmes, S. (PI)

Filter Results: