Building Your Tree Interactively with CHAID

IBM SPSS MODELER ANALYSIS

Need of segmentation

Problem:

10,000 Customers - we know their age, city name, income, employment status, marital status etc.

You have to sell iPhones (each costs $1000) to the people in this group.

How to go about selling the products.

Need of segmentation

Solution

Divide the whole population into two groups employed / unemployed

Further divide the employed population into two groups high/low salary

Further divide that group into having children

10000 customers

Unemployed 3000

Employed7000

High Income5000

Low Income2000

Children1800

No Children200

Decision Tree Decision tree is a decision support tool

One way to display an algorithm.

Decision Tree Vocabulary◦ Drawn top-to-bottom or left-to-right

◦ Top (or left-most) node = Root Node

◦ Descendent node(s) = Child Node(s)

◦ Bottom (or right-most) node(s) = Leaf Node(s)

◦ Unique path from root to each leaf = Rule

Root

Child Child Leaf

LeafChild

Leaf

Decision Tree Algorithms

Decision Trees Algorithm – Answers?

(1) Which attribute to start?

(2)Which Split to consider?

(3) Which attribute to proceed with?

(4) When to stop/ come to conclusion?

Decision Tree Algorithms

Decision Tree Algorithms◦ Hunt’s Algorithm (one of the earliest)◦ CART◦ ID3◦ C4.5◦ SLIQ◦ SPRINT◦ CHAID

CHAID OVERVIEW

Chi-square Automatic Interaction Detector (CHAID) 

Discovers the relationship between variables.

It builds a predictive model or tree

Data use: ◦ Nominal ◦ Ordinal◦ Continuous splits in categories

ex. If there are 1000 samples, creating 10 equal population bins would result in 10 bins, each containing 100 samples.

The decision or split made at each node is still based on a single variable, but can result in multiple branchesThe split search algorithm is designed for categorical variables.

Build a Tree Interactively with CHAID to Predict

Churn

DEMO

Import and instantiate the Data

Add and configure the CHAID node

Grow Branch with Custom Split

Select Predictors

Define Split

Evaluate the findings

TASKS

CHAID ON SPSS Modeler

Selecting variables

Summary