Association_Rule - Association Rules and Market Basket...

Info iconThis preview shows pages 1–12. Sign up to view the full content.

View Full Document Right Arrow Icon
3/31/2009 1 Association Rules and Market Basket Analysis Kwok-Leung Tsui Industrial & Systems Engineering Georgia Institute of Technology
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
3/31/2009 2 • Supervised learning: – Learning with a teacher – Classification, e.g. online shoppers (buyers Vs. non-buyers) • Unsupervised learning: – Learning without a teacher – Clustering, e.g. online shoppers (segmentation of non- buyers) • Other related terms: – Machine Learning (analogies to human receiving) – Neural Networks (biological analogies to brain) Supervised & Unsupervised Learning
Background image of page 2
3/31/2009 3 Inputs: (Predictors, independent variables, y) – A set of variables which are measured or preset. Outputs: (Responses, dependent variables, x) – A set of measurable variables which are influenced by the inputs Steps: – Establish models / systems(y hat) based on collected inputs & outputs (x and y). – Predict the values of outputs based on the established models / systems and a new set of specified inputs. Supervised Learning
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
3/31/2009 4 • Learning without a teacher • Statistical Definition – Observe N vectors from the population distribution – Directly inference on the properties (e.g. relationship, grouping) on the population distribution • Dimension of the observation (# of variables or attributes) is often very high (much higher than that in supervised learning) • No clear measure of success – The success is often judged (subjectively) by the value of discovery knowledge or the effectiveness of the algorithm Unsupervised Learning
Background image of page 4
3/31/2009 5 Association Rule : – Single-Dimensional Association Rule from Transaction Database – Multi-Level Association Rule from Transaction Database – Multi-Dimensional Association Rule from Relational Data Base and Data Warehouse – Correlation Analysis Problems in Unsupervised Learning
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
3/31/2009 6 Examples : – Multi-Level Association Rule : • Computer: desktop (IBM, Dell), laptop (Toshiba, Sony) • Software: educational (Microsoft, …), financial (…, …) • Printer: color (HP, Epson), B/W (HP, Sony) • Rule e.g.: {IBM desktop computer => B/W printer} – Multi-Dimensional Association Rule : • buys(X, “IBM desktop computer”) => buys(X, “Sony B/W printer”) • Age(X, “20 to 29”)& Occupation(X,”students”) => buys(X, “laptops”) Association Rules
Background image of page 6
3/31/2009 7 Clustering – Partitioning methods: • K-means, K-medoids – Hierarchical Methods: • BIRCH, CURE, chameleon, algorithms – Density-Based Methods: • DB SCAN, OPTICS, DENCLUS – Grid-based methods: • STING, Wave cluster, CLIQUE – Model-Based Clustering: • CoBWEB (tree-model) • Neural Network model Algorithms in Unsupervised Learning
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
3/31/2009 8 Association Rules – Market basket analysis – Generalized association rules Cluster Analysis – K-mean algorithms – Clustering algorithms – Combinatorial algorithms Other Multivariate Methods – Principle components – Factor analysis and latent variables – Projection pursuit – Multi-dimensional scaling Models for Unsupervised Learning
Background image of page 8
3/31/2009 9 Learn a method for predicting the instance class from pre labeled (classified) instances Classification Example Classification Models
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
3/31/2009 10 Find “natural” grouping of data given un labeled data Clustering Example
Background image of page 10
3/31/2009 11 Market Basket Analysis
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 12
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 47

Association_Rule - Association Rules and Market Basket...

This preview shows document pages 1 - 12. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online