Statistical Methods for Large, Complex Data

Provided by: 

S.ANNA

From: 

Altro PhD (Economics)

Lecturers: 

Francesca CHIAROMONTE

Semester: 

2

Hours: 

10

Exam: 

N

Educational Goals: 

Outline: Lecture 1: Computational Assessment of Statistical Procedures. Resampling (e.g. Jacknife, Boostrap), Cross-Validation schemes and their uses. Lecture 2: High Dimensional Supervised Problems. Linear and Generalized Linear Models (review, including traditional feature selection), Shrinkage and Sparsification (Ridge, LASSO and other developments). Lecture 3: Ultra-High Dimensional Supervised Problems. Feature Screening algorithms for linear and generalized linear models, and model-free. Lecture 4: Ultra-High Sample Sizes. Significance and Effect Sizes, Subsampling strategies for Big Data.
Zircon - This is a contributing Drupal Theme
Design by WeebPal.