STAT 348
Download as PDF
Predictive Analytics
Course Description
Introduction to the predictive analytic workflow including pre-processing data, feature engineering, model building and analysis. Models covered include regression, logistic regression, trees, random forests, penalized regression, K-nearest neighbors, support vector machines, multi-layer perceptrons and time series models. While not-exclusively for actuaries, this class will help actuarial science students prepare for Exam SRM and CAS's Exam MAS-I.
When Taught
Fall
Fixed/Max
3
Fixed
3
Fixed
0
Title
Basics of Statistical Learning
Learning Outcome
Perform exploratory analysis and understand how to choose what type of model to use for particular characteristics of the data and the context of a specific business problem
Title
Generalized Linear Models
Learning Outcome
Understand generalized linear models and how to select distributions and link functions. Estimate parameters, perform diagnostic tests, interpret model output, and perform prediction
Title
Time Series Models
Learning Outcome
Understand basic concepts of stochastic processes. Describe and use common time series models.
Title
Decision Trees
Learning Outcome
Explain, interpret, and use regression trees for regression and classification
Title
Principal Component Analysis
Learning Outcome
Calculate and interpret principal components for a data set
Title
Cluster Analysis
Learning Outcome
Cluster observations using K-means and hierarchical clustering
