MGT/B 296 Business Intelligence Technologies Data Mining Spring 2010

Transcription

1 MGT/B 296 Business Intelligence Technologies Data Mining Spring 2010 University of California, Davis Graduate School of Management Professor Yinghui (Catherine) Yang Room 3418, Gallagher Hall, UC Davis Davis: Tuesday 12:10pm 3:00pm, 2310 Gallagher Hall San Ramon: Saturdays 9-12:00n and 1-4:00pm on Apr. 3, 17, May 1, 15, 29 Course Description Data is a key source of intelligence and competitive advantage for business organizations. With the explosion of electronic data available to organizations and the demand for better and faster decisions, the role of data driven intelligence is becoming central in organizations. Data mining is the process of converting the raw data into useful knowledge required to support decision-making. It automates the process of knowledge discovery, making us orders of magnitude more productive in our search for useful information than we would be otherwise. It also increases the confidence with which we can make business decisions. Virtually every business organization these days is in the process of exploring and implementing data mining solutions to core business problems. This course is essential for anyone interested in understanding how to get the maximum value from data, especially when abundant data are available. Application areas covered in this course include marketing, customer relationship management, financial forecasting, risk management, personalization, Web searching, etc. The course focuses on two subjects simultaneously: 1. The essential data mining and knowledge representation techniques used to extract intelligence from data and experts. Such techniques include decision trees, association rule discovery, clustering, classification, neural networks, nearest neighbor, link analysis, etc. 2. Common problems from Marketing, Finance, and Operations that demonstrate the use of various techniques. The course is structured so that it is suitable both for students interested in a conceptual understanding of data mining and its potential as well as those interested in understanding the details and acquiring hands-on skills.

2 Intended Audience and Prerequisites The course is recommended for students interested in understanding the techniques and applications of data mining for making intelligent business decisions in data-rich organizations. Students interested in Marketing or Finance will also find this course useful. No prior knowledge is required for taking this course. MGT/MGP 207 or MGT/MGP 287 are not prerequisites of this course. Textbook: Software: Data Mining Techniques: For Marketing, Sales, and Customer Relationship Management, Second Edition Michael Berry and Gordon Linoff, 2004, Wiley, ISBN-10: ISBN-13: We will use a free data mining software WEKA for our data mining needs. No programming skill is needed. Course Web site: We will make extensive use of the Web site in this course. Important information for the class (e.g., announcements, lecture notes and other handouts) will appear on the Web site. You should make a habit of checking it regularly. Note: This syllabus may change during the quarter. The course Web site will always have the most up-to-date syllabus. Office Hours: Knock on my door, make an appointment, or just give me a call. Grading: Components Grades Class Participation 15% Homework 50% Term Project 35%

3 Participation: Make sure you read the cases carefully before each class, and actively participate in discussion and offer insightful comments in class. The following scheme is used to grade student participation is each class: 3: Excellent 2: Good 1: OK 0: Absent with good reason and advance notification -3: Absent with no reason Term Project: Each group (2 or 3 students) is responsible for choosing a company to study its data utilization situation and make recommendations about how its data should be used for generating business intelligence. There are two phases. In phase 1, which is due in the middle of the quarter, each group needs to turn in a 2-3 page write-up about the company they picked and generally describe its data situation. Phase 2 is due before the last class, in which each group presents its project to the entire class. In the final project report, each group needs to follow detailed guidelines to analyze the company s data strategies, and make detailed recommendations based on the techniques and strategies we learned from the class. If possible, a group can also obtain some sample data from the company, and make detailed analysis for the company. Homework: There are 2 homework assignments in total. Each counts 25% of the total grades. Each homework assignment may include problem solving, data analysis and case discussion. You should work on them individually. Policies: I ll try my best to create a healthy learning environment both in the classroom and after class. Non-class related activities are discouraged in class. Please try your best to be on time for the class. After a class, you are responsible for reviewing the materials covered and reading the related text before the next class. Attendance: Attendance is required for this course.

4 Class Schedule: (Subject to change) Davis: Week Topic 1 (Mar 30) Course Overview, Intro to Data Mining 2 (Apr 6) Market Basket Analysis & Association Rules, CRM 3 (Apr 13) Market Segmentation & Clustering, Prepare data 4 (Apr 20) Prediction & Classification Decision Tree 5 (Apr 27) Personalization & Nearest Neighbor 6 (May 4) Financial Forecasting & Neural Networks 7 (May 11) Link Analysis & Web mining 8 (May 18) Misc topics 9 (May 25) Memorial Day, Guest lecture scheduled on a different day in San Ramon 10 (June 1) Term project presentations San Ramon: Week Topic 1 (Apr 3) Course Overview, Intro to Data Mining 2 (Apr 3) Market Basket Analysis & Association Rules, CRM 3 (Apr 17) Market Segmentation & Clustering, Prepare data 4 (Apr 17) Prediction & Classification Decision Tree 5 (May 1) Personalization & Nearest Neighbor 6 (May 1) Financial Forecasting & Neural Networks 7 (May 15) Link Analysis & Web mining 8 (May 15) Guest Lecture 9 (May 29) Misc topics 10 (May 29) Term project presentations * Guest speaker date may change according to guest speaker s schedule. Due Dates: Davis San Ramon Group Decision April 6 April 14 Project Phase 1 April 20 April 28 Homework 1 April 27 May 5 Homework 2 May 18 May 19 Final Project June 1 May 29

5 Reading list: 1. Course Overview, Introduction to Data Mining A Case Study in Business Data Mining Bank of America (textbook pp ) A Wireless Communication Company Makes the Right Connections (textbook pp ) Neural Networks and Decision Trees Drive SUV Sales (textbook pp ) Textbook: Chapter 1, 2, 3 2. Market Basket Analysis, Association Rules How to Increase Sales in Retail with Market Basket Analysis (on course website) Textbook: Chapter 4, 9 3. Market Segmentation, Clustering Clustering Town The Boston Globe (textbook pp ) Textbook: Chapter 11, Predication, Classification, Decision Tree Fleet Uses CART Data Mining Technology to Understand Customer Characteristics and Habits (on course website) Molecular Data Mining Tool: Advances in HIV Research (on course website) Textbook: Chapter 6 5. Personalization, k-nearest Neighbors Firefly Networks (HBS case, available from Study.net) Polyphonic HMI: Mixing Music and Math (HBS case, available from Study.net) Textbook: Chapter 8, Financial Prediction, Neural Networks A Neural Approach to the Market (on course website)

6 Finding Clusters A Large Bank (textbook pp ) Textbook: Chapter 7 7. Link Analysis, Web Mining, Search Engine Marketing MedNet.com Confronts Click-Through Competitionet (HBS case, available from Study.net) Google Advertising (HBS case, available from Study.net) Textbook: Chapter Misc. Topics Topics to be decided based on students interests and class progress. Textbook: Chapter 14, 16, 18

7 Term Project The term project is intended to provide you with valuable opportunity to be involved with a company s data-driven decisions. The main goal is to thoroughly understand how a company leverages its data for business intelligence, and to connect what you learn from this class with the real problems in order to make valuable recommendations to the company. The project contains the following steps. Step 1: Form a group The size of the group cannot exceed 3 students. You have a week from the first class to form the group, so start early. Please me your decision by the due date. Step 2: Identify a company to study This company can be your previous employer, current employer or any other company that grants you access to their information. Step 3: Turn in Phase 1 report Each group needs to turn in a 2-3 page write-up about the company they picked and generally describe its data situation. By the time you turn in your phase 1 report, you should have a rough idea about the amount and depth of the information the company allows you to access, and please describe that in the report too. Step 4: Turn in the final project report When writing your final report, try to touch upon the 11 steps of data mining methodology in Chapter 3 of the textbook, or use these steps to structure your report if necessary. Feel free to structure your report differently if you feel it can help you bring out the insights better. You can decide to focus more on the strategy side of data mining or technical side of data mining. When analyzing the current situation of the company, please discuss the strength and weakness of its current practice, and how business intelligence is derived from data. When making recommendations, please try to back up your recommendations as much as possible. The report can also include information such as the software used, data flow within the organization, and other valuable observations. If you are able to obtain the company s data and have interests in analyzing them, please try to connect the results of the data analysis with strategy-level recommendations. This is optional. Step 5: Class Presentation Your presentation should at least include the companies background, how its data is handled, how business intelligence is generated from the data and your recommendations.

8 Every member of the group should try to be involved in the presentation. Please allow minutes for the presentation, during which I will ask questions. Grading: The term project in total counts 35% of the total grade. Phase 1 report: 5% Final report: 25% Class presentation: 5%