Data Mining Applications in Business by uyk11591

VIEWS: 0 PAGES: 7

More Info
									                         Data Mining for Business
                                               Course Syllabus

Course Number:
Class Number:
Classroom:
Class Hours:
Term:
Instructor:
Office:
E-mail:
Phone:

Course Description
The last 10 years have seen an explosion in the quantity of data available to businesses. Transactional
data from point-of-sale scanners are now routinely available; data from direct marketing is growing
exponentially; and e-commerce and web-browsing/click stream data has recently become available.
There is strong interest in extracting value from this data. This course discusses how data mining
technologies are used to transform large quantities of data into information to support tactical and
strategic business decisions. We discuss applications of data mining technologies in customer
relationship management (CRM), direct marketing, e-commerce, finance, and retailing. Although we
discuss some workings of the technologies, the focus of the course is learning when and how to use the
technologies in business applications. We approach the material from the perspective of a business
analyst. The course is designed primarily for students concentrating in information systems, management
science and statistics, e-commerce, and marketing.

Prerequisites
It will be assumed that you have a good understanding of data, models, and decisions. It is also helpful if you have
completed coursework on Data Management Systems or are taking this coursework concurrently.

Required Texts
Dhar & Stein. Seven Methods for Transforming Corporate Data into Business Intelligence. Prentice Hall,
1997. (D&S)

Course packet (description of materials included in this syllabus)

Required Software
Clementine GradPack from SPSS and SPSS’ Amos software (available in the computer lab or individual
copies may be purchased through the bookstore.

Learning Objectives
1. Learn how businesses can gain competitive advantage through the warehousing and mining of data.
2. Enhance your ability to summarize, visualize, and analyze data in support of managerial decision
   making.
3. Become proficient in the use of data mining technologies in business applications.

Instructional Methods Used
The course uses lectures, discussions of examples/applications, assigned readings, exercises, team
projects, and two in-class exams to achieve the above learning objectives.
Class Meetings and Preparation
Class meetings consist of lectures, discussions of examples/applications, and software demonstrations.
Please arrive to class on time so as to avoid disruption.

The assigned readings for each week are listed in the Schedule provided near the end of this syllabus.
You are expected to complete these readings in advance of each class. You are also expected to
participate in classroom discussion. Class participation does not solely mean responding to questions:
you are also urged to ask questions that lead to fruitful discussion, provide comments on other students’
comments/responses, and offer insights based on your experience and background. Class participation
requires you to listen carefully and respectfully to others’ comments and opinions.

You are strongly encouraged to bring any interesting articles/reports on data mining to the attention of the
instructor. Such material may very well be incorporated into class discussion.

Software and Computing
An important part of the course is the use of contemporary software for analysis and management of data
sets. Because many topics and concepts in data mining are learned most efficiently through hands-on
work, we will spend a fair amount of time (both during and outside of class) analyzing and mining
business data with contemporary software. The goal is not to master all the features and intricacies of
specific software packages (which you may well make little to no use of again) but rather to gain a better
understanding of (a) how data mining is applied and (b) what is involved in data mining projects (including
the steps of the data mining process).

The primary software packages used in the course are Clementine (www.spss.com ) and AnswerTree
(www.spss.com ). Both packages are available in the computer lab or individual copies may be
purchased through the bookstore.

The use of software will be demonstrated during class meetings. You are required to become familiar with
the aspects of the software covered in class (including interpretation of output).

You are NOT required to purchase manuals for the software used in the course, although you are
welcome to do so. Instead of purchasing manuals, you may refer to the online help provided with the
software as well as the documentation (in pdf format) for Clementine and AnswerTree provided on the
computers in the labs. Furthermore, saving the Clementine and AnswerTree analyses done during the
class meetings will help you learn the software by documenting the settings and steps used in analyses.

Exercises
You will be assigned exercises periodically during the semester. Some of these exercises will be
collected and graded. Each exercise is to be done on your own—not as part of a group. Although you
may seek assistance from others in running and executing software, you may neither give nor receive
help in other aspects of the exercises. No late exercises can be accepted.

Team Projects
You will be assigned two team projects during the semester. Each project will involve mining/analyzing a
business data set. The instructor will assign the teams for the projects. Each team will be comprised of
three students. The due dates for the two projects are listed in the Schedule provided near the end of this
syllabus. No late projects will be accepted. Each team will make a 10-minute presentation on Team
Project 2 during final exam period. A complete set of project guidelines and requirements will be issued



Prepared by:
Paul Zantek, Ph.D
Assistant Professor, Decision and Information Technologies
The University of Maryland, College Park
Course: BUDT 736
during the semester. Your grade for each of the projects will be determined through assessment of the (a)
analysis, (b) written report, (c) peer evaluation, and (d) oral presentation (in the case of Team Project 2).


Examinations
Two midterm examinations will be given. Please see the Schedule near the end of this syllabus for the
dates of these exams. You must be prepared to present your official University-issued photo ID at each of
these exams. Absence from an exam will result in a score of zero for the exam unless prior arrangements
have been made with the instructor.

Grading Policy
Your final grade for the course will be based on your performance on the exercises, the two exams, and
the two team projects. The weights given to each of these components are as follows.
     Individual Exercises 15%
     Team Project 1 15%
     Team Project 2 15%
     Test 1 25%
     Test 2 30%
You are expected to keep all course materials (e.g., syllabus, handouts, notes, assignments, and graded
exams) until you retrieve your final course grade at the end of the semester. Please keep backup copies
of all assignments that you turn in, as well as all graded assignments that have been returned to you.
These back-up copies will be needed in the unlikely event that a grading error is made.

Attendance
You are expected to attend each class. Frequent absence has a negative effect on your learning as well
as your course grade. If you miss a class session, you are responsible for all the material covered in that
session. Thus, you are urged to ask one or two other students in the class for a copy of their notes,
handouts, etc. from the missed session.




Prepared by:
Paul Zantek, Ph.D
Assistant Professor, Decision and Information Technologies
The University of Maryland, College Park
Course: BUDT 736
Course Schedule
 Class     Dates       Topics                                    Readings and Exercises
                                                                 D&S: Dhar and Stein
                                                                 CP: Course Packet
                       Definition of Data Mining                 D&S: Chapters 1-3
 1
                                                                 CP: “Selling is Getting Personal”
                       Motivation for Data Mining                CP: Shearer, “The CRISP-DM
                                                                 Model: The New Blueprint for Data
                       Strategic Business Applications of Data   Mining” CRISP-DM: Chapter 1
                       Mining                                    CP: “HSBC Bank USA”

                       Steps in Data Mining

                       Continuation of Lecture 1
 2
                       Market Basket Analysis                    CP: Berry & Linoff, “Market Basket
 3
                                                                 Analysis”
                       Association Rules                         CP: Holstein, “Data-Crunching
                                                                 Santa”
                       Chi-Squared Test of Association           CP: Streitfeld, “Who’s Reading
                                                                 What?”

                       Continuation of Lecture 2
 4
                       Homework:
                       Assignment 1

                       Continuation of Lecture 2
 5

                       Continuation of Lecture 2
 6
                       Homework:
                       Assignment 2

                       Targeting of Customers in Direct          D&S: Chapter 10
 7
                       Marketing                                 CP: Headden, “The Junk Mail
                                                                 Deluge”
                       Model/Method: Classification Trees        CP: “British Telecommunications”




Prepared by:
Paul Zantek, Ph.D
Assistant Professor, Decision and Information Technologies
The University of Maryland, College Park
Course: BUDT 736
                       Continuation of Lecture 3
 8
                       Due:
                       Assignment 1

                       Continuation of Lecture 3               CP: Binkowski & Rosenwein,
 9
                                                               “Prescription for Online Pharmacy
                       Homework:                               Success”
                       Begin work on Team Project 1

                       Continuation of Lecture 3
 10
                       Due:
                       Assignment 2

                       Customer Retention and Analysis of      D&S: “Winterthur Insurance”
 11
                       Customer Attrition (“Churn Analysis”)

                       Artificial Neural Networks

                       Classification Trees

                       Continuation of Lecture 4
 12

                       Test 1
 13

                       Lab Session to Work with Software
 14

                       Continuation of Lecture 4               D&S: “Financial Market Analysis
 15
                                                               and Prediction”, pp. 228-236
                       Continuation of Lecture 4
 16
                       Due:
                       Team Project 1

                       Clustering and Segmentation             CP: Thomas, “Getting to Know
 17
                                                               You.com”
                       Kohonen Networks

                       Homework:
                       Begin work on Team Project 2
                       Continuation of Lecture 5
 18



Prepared by:
Paul Zantek, Ph.D
Assistant Professor, Decision and Information Technologies
The University of Maryland, College Park
Course: BUDT 736
                       Continuation of Lecture 5
 19

                       Clustering and Segmentation
 20
                       K-Means Algorithms

                       Homework:
                       Assignment 3

                       Continuation of Lecture 6
 21

                       The Data Mining Process                       REVIEW CP: Shearer, “The
 22
                                                                     CRISP-DM Model: The New
                       Preparing Data for Mining                     Blueprint for Data Mining” CRISP-
                                                                     DM: Chapter 1

                       Continuation of Lecture 7
 23
                       Due:
                       Assignment 3
                       A Recent Topic or Issue in Data Mining
 24
                       (topic and readings to be determined)

                       Test 2 (a cumulative exam)
 25

                       First Union National Bank Case Study (in
 26
                       D&S)

                       Homework:
                           Prepare for discussion of the case
                            study “Managing Credit Risk Through
                            Embedded Intelligence in On Line
                            Transaction Processing: First Union
                            National Bank, Charlotte, NC” by Dhar
                           Turn in a short paper addressing the
                            following:
                              o Define the problem (briefly).
                              o Use the framework in Chapter 3
                                   of D&S to articulate FUNB’s
                                   specific needs/requirements.
                              o What solutions (tools,
                                   algorithms, methods, etc.)
                                   discussed in the course seem to
                                   meet FUNB’s requirements?
                                   Which of them seem best suited
                                   to FUNB’s problem?


Prepared by:
Paul Zantek, Ph.D
Assistant Professor, Decision and Information Technologies
The University of Maryland, College Park
Course: BUDT 736
                       Continue Discussing First Union National
 27
                       Bank Case Study

                       No Class- Reading Week
 28

                       No Class- Reading Week
 29

                       No Class- Reading Week
 30

                       Team Project 2 presentation and written
 31
                       report due



References (all are provided in the course packet)

Berry & Linoff, Data Mining Techniques for Marketing, Sales, and Customer Support, Wiley, 1997, pp.
      124-156.

Binkowski, E. S., and Rosenwein, M. B. “Prescription for Online Pharmacy Success,” OR/MS Today,
     August 2000 (pp. 46-50).

“British Telecommunications,” article at http://www.spss.com/success/pdf/BTAPP-0899.PDF

Dhar, Vasant, “Managing Credit Risk Through Embedded Intelligence in On-Line Transaction Processing:
      First Union National Bank, Charlotte, NC”

Headden, S., “The Junk Mail Deluge,” U.S. News and World Report, December 8, 1997 (pp. 40-48).

Holstein, W., “Data-Crunching Santa,” U.S. News and World Report, December 21, 1998 (pp. 44-48).

“HSBC Bank USA,” article at http://www.spss.com/success/pdf/HSB-0200.pdf

“Selling is Getting Personal,” Consumer Reports, November 2000 (pp. 16-20).

Shearer, C., “The CRISP-DM Model: The New Blueprint for Data Mining,” Journal of Data Warehousing, 5
     (4), pp. 13-22.

Streitfeld, D., “Who’s Reading What?,” Washington Post, August 27, 1999.

Thomas, S. G., “Getting to Know You.com,” U.S. News and World Report, November 15, 1999 (pp. 102-
    111).

“Winterthur Insurance,” article at http://www.spss.com/success/pdf/WINAPP-0200.PDF




Prepared by:
Paul Zantek, Ph.D
Assistant Professor, Decision and Information Technologies
The University of Maryland, College Park
Course: BUDT 736

								
To top