COURSE NUMBER:  MBA 247.11

The course is dual-listed with the Evening-Weekend MBA Program

COURSE TITLE:  Descriptive and Predictive Data Mining

INSTRUCTOR:  Andy Shogan

E-MAIL ADDRESS:  andy@haas.berkeley.edu

PREREQUISITE(S):  Knowledge of the basics of Excel AND completion of MBA/EWMBA200S “Data and Decisions”.  Building on the core course, this course places the core course’s statistical models (e.g., regression) into the larger context of Business Analytics.

CAREER FIELD:  Data Mining is relevant to all career fields.

CLASS FORMAT:  A mixture of lecture (70%) and in-class group exercises (30%).

REQUIRED READINGS:  This course covers Chapters 1, 3, 4 & 9 from the textbook Essentials of Business Analytics (3rd edition), by Camm, Cochran, Fry, Ohlmann, Anderson, Sweeney, & Williams, published in 2018 by Cengage in Learning.

BASIS FOR FINAL GRADE:  3 Groups Assignments and an Individual Take-Home Final Exam on Sunday, 10/20.

ABSTRACT OF COURSE'S CONTENT AND OBJECTIVES:  Every hour of every day, you either use data or create data for someone else to use.

• How does your e-mail server identify spam?
• How does Amazon customize the books it recommends to you, how does Netflix customize the movies it recommends to you, and how does a web site customize the ads it displays to you?
• How does your grocery store know which coupons to include with your receipt?
• How does a bank know whether to approve or reject a mortgage application?
• How does a company’s HR Director predict whether a job candidate is likely to be a long-term employee, or how does a company (e.g., American Airlines or Verizon) predict whether a customer is considering switching to a competitor?
• How does the IRS decide whether to audit a tax return?
• How does a doctor use data to better diagnose you?
• How does a sports team’s coach/manager use data to make real-time decisions during the game/match?

The above are just a few questions that you will be able to answer after enrolling in this course that focuses on analytical approaches to data-driven decisions.  This course covers 3 distinct topics, using both native Excel functions and an add-in to Excel called Analytic Solver Data Mining. (You can use this add-in free-of-charge either virtually from the Haas Terminal Server or physically in the computer classroom S300T, AND/OR you can install it on your personal computer for 140 days for about $25).

• Data Visualization (as the first step to interpret data, and as the last step to convey data)
• Excel’s PivotTable & PivotChart
• Excel’s Charts (Scatter, Column/Bar, Line, Heat-Map, Sparkline, Bubble, & 3D-Map)
• Descriptive Data Mining (aka Unsupervised Machine Learning) – using data to identify relationships
• Cluster Analysis via k-Means Clustering
• Association Rules
• Predictive Data Mining (aka Supervised Machine Learning) – using data to make a prediction
• Data Partitioning (to prepare for a Predictive Model)
• Measuring the Accuracy of a Predictive Model for a 2-Category Outcome
• Confusion Matrix
• Error Rates (Class 1, Class 0, & Overall)
• The Tradeoff between Class 1 Error Rate and Class 0 Error Rate
• Two Predictive Models for a 2-Category Outcome
• k-Nearest Neighbors
• Classification & Regression Trees (CART)

After completing this course, you will understand the basic concepts and techniques of data-driven decisions, thereby having the capability not only to interact meaningfully with experts but also to perform the initial analysis yourself and ask critical questions.

RELATIONSHIP TO EWMBA 296.9 “DATA SCIENCE” (offered by Greg La Blanc ) Although there will be some overlap between the two courses, you are permitted to enroll in both courses in either order.  The primary difference between the two courses (other than 2 units versus 1 unit) is that course approaches Data Mining with a top-down strategic focus, whereas this course approaches Data Mining with a bottom-up analytical focus.

BIOGRAPHICAL SKETCH:  Andy SHOGAN joined the Haas School’s faculty after receiving an A.B. in Mathematics from Princeton and a Ph.D. in Operations Research from Stanford.  At the Haas School, Andy served several times as chair of the Operations & Information Technology Group, and, during 1991-2007, he served as the Haas School’s Associate Dean for Instruction.  In addition to teaching at Haas, Andy also performs educational-consulting projects for Haas (most recently in Vietnam, Saudi Arabia, & Sri Lanka).  In 2007, Andy was awarded the Chancellor’s Distinguished Service Award for his long-time service to both Haas and the University.  For his teaching, Andy has twice received the Haas School’s Cheit Award for Teaching Excellence (once from the Full-Time MBA students and once from the EWMBA students), and he has received the campus-wide Distinguished Teaching Award from the University’s faculty.  In addition to being a Haas faculty member, Andy served for 10 years as a Visiting Professor at Switzerland’s Lorange Institute of Business (formerly GSBA Zurich) and is now a Visiting Professor at Denmark’s AVT Business School.   Andy has taught a variety of executive education programs in the United States, China, Thailand, Taiwan, Jamaica, Mexico, and Switzerland.  In 1988, his textbook Management Science was published by Prentice-Hall.  Andy and his wife, both natives of Pittsburgh, PA, now reside in Orinda, CA, and have 3 sons, 2 daughters-in-law, 2 grandsons, and 1 granddaughter.