Вы находитесь на странице: 1из 14

CART (Data Mining Software)

Sabam Parjuangan Atma Hadiansyah Masrizal Sahara Abdi

Overview

What is the Cart? Define the goal of Cart Simple Case using Cart

What is the Cart?

Cart is software a robust decision-tree tool for data mining, predictive modeling, and data processing. CART can be applied to data from any subject. We have come across CART models in agriculture, banking, genetics, mark eting, security, and zoology, among many others, and the citations to CART number thousands. in the

About CART Menus

When you first start CART you see one set of menus but the menu items
will change as you progress through an analysis. Menus can change to reflect the stage of your analysis and the window you have ac tive. As a result, not all menus are always available. Similarly, when not accessible, the commands that appear in the pull-down menus and the toolbar icons are disabled.

FILE Open data set, Navigator file, Grove File, or command file Save analysis results, Navigator file, Grove file, or command file Open a CART notepad for creating command scripts Specify printing parameters Activate interactive c ommand mode Submit batch command files EDIT Cut, copy and paste s elected text Specify colors and fonts Control reporting options

VIEW Open command log View data View descriptive statistics Display next pruning Assign class names and apply colors View main tree and/or sub-tree rules Overlay gains charts Specify level of detail displayed in tree nodes EXPLORE Generate frequency distributions MODEL Specify model setup parameters Grow trees/committee of experts Generate predictions/score data Translate models into SAS , C, or PMML TREE Prune/grow tree one level View optimal/minimum cost/maximal tree View tree summary reports REPORT Control CART reporting facility

Set random number seed


Specify default directories

Selecting Target and Predictor Variables

For this analysis, the binary categorical variable TARGET (coded 0/1) is the target (or dependent) the variable. To mark the target variable, use the to scroll down the variable list until the TARGET name is visible and place a check mark as shown below.

Timeline

Timeline

Date 1

Date 2

Date 3

Date 4

Timeline

Date 1

Date 3

Date 2

Date 4

Looking Ahead

When is the next milestone? What are the expected deliverables? Known risks and issues
o

What is the investigation timeline for these issues?

What are the immediate next steps?

Dependencies and Resources


Vendors

Remote Teams

Manufacturing

Project

Engineering

Sales

Appendix

Appendix

Budget Design documents Marketing plan Supplemental documents Contact information

Вам также может понравиться