BI Tools & Techniques Data Mining Assignment
In this assignment you will choose to use a free/open-source data mining tool, KNIME. You are to analyze a given dataset (about the voting behavior of a number of counties in the U.S.) to develop and compare at least three different types of prediction (i.e., classification) methods that predicts weather a county will say "yes" or "no" to legalizing gaming at the ballot). Here are the specifics for this assignment:
Use the following tools - KNIME (download and install on a PC/Laptop).
Download "Voting Behavior" data and the brief data description from the D2L - The data is given in MS Excel format.
Follow the 6 steps in CRISP-DM process model
- Understand the domain and the problem you are trying to solve (via literature).
- Understand, and preprocess the data (be very critical about the data).
- Develop at least three classification models (e.g., Decision Tree, Logit, ANN, etc.).
- Compare the accuracy results (use confusion matrixes and comment on the outcome).
Present your results in an organized report
- Include a cover page.
- Write an "Executive Summary" (1 page long).
- Use the 6 steps in CRISP-DM to organize the remainder of the report.
- Include a conclusion page, where you need to comment on the tool and techniques you've used. What was good and what was bad, etc.
- Make sure to integrate figures (graphs, charts, tables, screen-shots) into the text as you see necessary. Do not use Appendixes.
- Try not to exceed 15 pages in total, including the cover (use 12 point Times New Roman fonts, and 1.5 line spacing).
Attachment:- Assignment Files.rar