Agenda Day Three

Event Home |  Agenda |  Registration Additional Information |  CPD |  Seminar Location |  Presenters

 

Agenda Day One | Agenda Day Two | Agenda Day Three

 

Multivariate Insights with Statistical Learning

Day three will conclude the seminar with an illustration of such “unsupervised learning” techniques as principal components analysis (PCA) and clustering, as well as an introduction to tree-based modeling. Each of these techniques can be used profitably in isolation or in conjunction with core techniques like GLM and regularized regression. Clustering and PCA will be applied to a rich dataset, and discussed as complementary unsupervised learning techniques that are useful for gaining insights into high-dimensional datasets. Tree-based modeling will be illustrated with the classification and regression trees (CART) algorithm, Random Forests, Boosted Trees, and the Multivariate Adaptive Regression Splines [MARS] algorithm. The topics will be covered at a conceptual level, with an emphasis on practical case studies.

Friday, December 14
7:30 a.m. – 8:30 a.m.
  • More realistic case study illustrating principal components and clustering analysis

Session Coordinator(s)

Facilitator(s)

8:30 a.m. – 9:30 a.m.
  • Recursive partitioning, splitting rules, measures of purity
  • “Pruning back” using cross-validation
  • Classification and Regression Trees [CART]
  • Illustrative case study
  • Session Coordinator(s)

    Facilitator(s)

    9:30 a.m. – 11:00 a.m.
    • Bagging and Random Forests
    • Boosted Trees
    • Multivariate Adaptive Regression Splines
    • Neural networks and deep learning (conceptual discussion)
    • Comparing multiple methods – machine learning case study

    Session Coordinator(s)

    Facilitator(s)

    11:00 a.m. – 11:30 a.m.
    • Q&A and open discussion

    Session Coordinator(s)

    Facilitator(s)