ISQS 7339 Business Analytics (Fall 2012)



The term project should apply coherently some of the advanced data mining techniques lectured in this semester:

1)    Explanatory use of data mining (interactions, intervening factors, precedents)

2)    Advanced decision tree applications

a.    Missing value handling

b.    Interactive decision tree

c.    Ensemble of trees

d.    Cross validation

3)    Advanced predictive modeling

a.    Input selections

b.    Generalized profit assessment

c.    Multi-target prediction

d.    Two-stage modeling

e.    Component modeling

4)    Market segmentation methods

5)    Text mining or sentiment analysis methods


Project Group:

A semester-long project is accomplished by the group up to two students (you can also do it individually). Each group will deliver a final report for business problem solving focusing on data preparation and data analysis.


Project Stage One (assignment due Oct 23)


Students are requested to submit a project proposal for further advices.


There are two choices of the project. You only need to do one of them


Type I Data mining skill-oriented


Report a preliminary data analysis project applying the decision tree approach. The report will contain the following contents:

1)    Application background of the data analysis project

2)    Business requirements of the project

3)    The strategies for problem solving

4)    Data set description

5)    Exploratory results of the data set

6)    Reference list


Extra SAS programming using Proc ARBOR, extensive data analysis, or elaborate data explanation with advanced concepts are encouraged.


Type II Academic research oriented


Develop a research paper proposal for the topic based on the application of decision tree. The completed proposal will cover the following contents:

1)    Research background

2)    Research problem(s)

3)    Relevant research work

4)    The methodology and data collection plan

5)    Theory development

6)    Expected research outcomes

7)    Reference list

Project Stage Two (during the week of Nov 5-9)


Project advisory meeting. Each project group will make an appointment for the meeting up to an hour.



Project Stage Three (assignment due Dec 9)


The improvements and extensions from the project in stage one. Alternatively, a new topic is also applicable. Additional requirements:


Type I Data mining skill-oriented

1)    Data analysis techniques applied and performance optimization, such as use of surrogate rule, imputation, etc. Discussions of the effects of different parameters are needed

2)    Show the applications of advanced data mining concepts and techniques, and, optionally, the optimization modeling.

3)    Well developed explanations of the data analysis

4)    Generalization of the outcomes, and the business value of the project

5)    The analysis outcomes in charts, and explanations

6)    The improvement of the SAS EM techniques

7)    SAS programming results and experiences


Type II Academic research oriented


1)    More specific literature review, covering both advanced techniques and the domain related focuses.

2)    Research methodology

3)    Insights in the context and good discussions of the results

4)    Demonstrative data analysis and data collection plan

5)    Expected publication opportunities


The final submission is an electronic file, the model in the XML form exported from SAS EM, SAS code (if any), and the dataset (if available).