Monday 31 December 2012

Decision Tree and Rule Base Segmentation


Rule induction is an important component of data mining, and this course covers two main styles of generating rules.

One style of machine learning is association learning. In association learning, the learning method searches for any association between features. That is, there is no specific target variable. An example is the recommendation systems used in many online shopping systems - If you bought X, then you may also like Y. We will look at the industry standard method: APRIORI.

A second style of machine learning is classification learning. In classification learning, a learning scheme takes a set of classified examples from which it is expected to learn a way of classifying unseen examples. These are forms of supervised learning, in which there is a specific target variable. We will look at two decision tree methods: C4.5 and CHAID. We will also look at some machine learning methods, such as PRISM and INDUCT.

Rule induction methods have a number of strong pluses going for them: They produce interpretable results; they are flexible and make no strong assumptions about model form; they perform well in practice.

Learn this in Anthony Babinec’s online course “Decision Tree and Rule Base Segmentation” at Statistics.com. For more details please visit at  http://www.statistics.com/decisiontrees/.  

Who Should Take This Course:
Analysts and researchers who need to know more about automated machine learning methods for generating association and decision rules: data miners, consultants, ecommerce analysts, market researchers, direct marketers, diagnosticians.

Course Program:

Course outline: The course is structured as follows
SESSION 1: Introduction
  • Overview of rule induction methods
  • APRIORI - Find all associations with at least a specified support, confidence, and number of elements

SESSION 2: Classification via CHAID
  • CHAID - Chi-square Automatic Interaction Detection - an exploratory method for large numbers of categorical variables
  • Response-based segmentation
  • Finding profitable segments
  • Understanding the quality of the CHAID solution

SESSION 3: Classification via C4.5 Decision Trees
  • C4.5 - A public-domain machine learning method
  • Decision trees versus rule sets
  • Limitations of tree-based methods
  • Successor methods to C4.5

SESSION 4: Rule Construction via Covering Algorithms
  • Using covering algorithms to construct rules
  • The PRISM rule learner
  • INDUCT - A modification of PRISM

Anthony Babinec is President of AB Analytics. For over two decades, Tony Babinec has specialized in the application of statistical and data mining methods to the solution of business problems. Before forming AB Analytics, Babinec was Director of Advanced Products Marketing at SPSS; he worked on the marketing of Clementine and introduced CHAID, neural nets and other advanced technologies to SPSS users. He is on the Board of Directors of the Chicago Chapter of the American Statistical Association, where he has held various offices including President.

This course takes place over the internet at the Institute for 4 weeks. During each course week, you participate at times of your own choosing - there are no set times when you must be online. Course participants will be given access to a private discussion board so that they will be able to ask questions and exchange comments with instructor, Dr. Daniel T. Kaplan. In class discussions led by the instructor, you can post questions, seek clarification, and interact with your fellow students and the instructor.

The course typically requires 15 hours per week. At the beginning of each week, you receive the relevant material, in addition to answers to exercises from the previous session. During the week, you are expected to go over the course materials, work through exercises, and submit answers. Discussion among participants is encouraged. The instructor will provide answers and comments, and at the end of the week, you will receive individual feedback on your homework answers.

For Indian participants statistics.com accepts registration for its courses at special prices in Indian Rupees through its partner, the Center for eLearning and Training (C-eLT), Pune.

For India Registration and pricing, please visit us at www.india.statistics.com.

For More details contact at
Call: 020 6600 9116

Websites:

Thursday 27 December 2012

Advanced Structural Equation Modeling


The American Psychological Association is debating whether to drop narcissism as a personality disorder, largely because it is hard to measure.  How do social scientists actually measure characteristics of personality?  Part of the answer is to measure more concrete attributes (e.g. via responses on a survey) and to treat, say, narcissism as a derived "latent variable." Causal models of relationships involving latent variables are the domain of structural equation modeling (SEM).  Randall Schumacker, author of a number of books in this area, will present his online course "Advanced Structural Equation Modeling" at statistics.com. For more details please visit at http://www.statistics.com/advancedsem.

Who can take this course:
Market researchers, educational researchers, sociologists and psychologists, political scientists, economists, and survey researchers.

Course Program:

Course outline: The course is structured as follows
SESSION 1: Multiple Indicators and Causes
  • Multiple Indicator and Multiple Causes (MIMIC) model
  • Multiple-Group model

SESSION 2: Multilevel Models
  • Multilevel (HLM) model
  • Mixture model
  • Structured Means model

SESSION 3: Multitrait, Multimethod, Interacties
  • Multitrait-Multimethod model
  • Second Order Factor model
  • Interaction models

SESSION 4: Latent Variable, Dynamic Factor
  • Latent Variable Growth Curve model
  • Dynamic Factor model
  • Power and Sample Size
  • Monte Carlo Methods

Dr. Schumacker is a professor at the University of Alabama and, in addition to "Advanced Structural Equation Modeling," he has co-authored "A Beginner's Guide to Structural Equation Modeling" (with Richard Lomax) and is the co-editor (with George Marcoulides) of "Advanced Structural Equation Modeling: Issues and Techniques and Interaction and Nonlinear Effects in Structural Equation Modeling." Dr. Schumacker was the founder, editor (1994-1998), and is the current emeritus editor of "Structural Equation Modeling: A Multidisciplinary Journal." He also founded the Structural Equation Modeling Special Interest Group at the American Educational Research Association.

This course takes place over the internet at the Institute for 4 weeks. During each course week, you participate at times of your own choosing - there are no set times when you must be online. Course participants will be given access to a private discussion board so that they will be able to ask questions and exchange comments with instructor, Dr. Randall Schumacker. In class discussions led by the instructor, you can post questions, seek clarification, and interact with your fellow students and the instructor.

The course typically requires 15 hours per week. At the beginning of each week, you receive the relevant material, in addition to answers to exercises from the previous session. During the week, you are expected to go over the course materials, work through exercises, and submit answers. Discussion among participants is encouraged. The instructor will provide answers and comments, and at the end of the week, you will receive individual feedback on your homework answers.

For Indian participants statistics.com accepts registration for its courses at special prices in Indian Rupees through its partner, the Center for eLearning and Training (C-eLT), Pune.

For India Registration and pricing, please visit us at www.india.statistics.com.

For More details contact at
Call: 020 6600 9116

Websites: