Week 3: Classification and Regression

Announcements

Tuesday

Recap last week

  • Highlight the patterns your recognized

  • Hands-on in-class exercises

  • Revisit Lab 2 requirements

Pattern Recognition

Narrow down to course introduction, Identify Patterns.

  • Classification

  • Regression

Textbooks

All textbooks are free available online, and are optinal, not required.

Algorithms

  • Linear Discriminant Analysis (LDA)

  • Support Vector Machine (SVM)

  • Nearest Neighbors

  • Decision Trees

  • Ensemble methods

Datasets

  • sklearn examples

  • 1. Supervised learning

  • 1.1. Linear Models

  • [the diabetes dataset]

  • 1.1.11. Logistic regression

  • Regularization path of L1- Logistic Regression, [Iris]

  • MNIST classification using multinomial logistic + L1, [MNIST]

  • 1.4. Support Vector Machines

  • SVM-Anova: SVM with univariate feature selection, [Iris]

  • 1.6. Nearest Neighbors

  • Nearest Neighbors Classification [Iris]

  • 1.10. Decision Trees

  • Plot the decision surface of decision trees trained on the iris dataset [Iris]

  • Understanding the decision tree structure [Iris]

  • 1.11. Ensemble methods

  • Plot the decision surfaces of ensembles of trees on the iris dataset [Iris]

  • 1.11.6. Voting Classifier [Iris]

  • 1.12. Multiclass and multioutput algorithms [Iris]

Code

In-class exercises:

*Explain to your classmates what is Classification
*Explain to your classmates what is Regression

Lab 03

Thursday

Supervised learning

  • From Algorithms perspective

  • From Dataset perspective

The Diabetes dataset

  • regression

  • Efron, Bradley, Trevor Hastie, Iain Johnstone, and Robert Tibshirani. "Least angle regression." The Annals of statistics 32, no. 2 (2004): 407-499.

  • Cited by 11138

  • Google Scholar has the Paper PDF file

  • lab 04

Midterm

  • research questions

  • literature review

  • paper format, two-page

  • timeline

  • team, student pairs

Math

Concept

  • loss function

  • Accuracy

  • Train-test

  • Cross validation

  • Overfitting and underfitting