1 COURSE SYLLABUS & OUTLINE Course Title: X 470.20 Predictive Analytics for Marketing, Reg#255343 Quarter SPRING 2015 Instructor: Meeting Dates: Time Ash Pahwa, Ph.D March 30 June 15, 2015 (Mondays) 7pm 10pm (*no meeting May 25 Memorial Day) Location: UCLA, Math Science building, Room 5128 Contact Information: Office Phone: (949) 378 1229 Email: ash@ashpahwa.com Website: www.ashpahwa.com Ash Pahwa, Ph.D., is an educator, author, entrepreneur, and technology visionary with three decades of industry and academic experience. He has founded several successful technology companies during his career, the latest of which is A+ Web Services (www.apluswebservices.com) which provides internet marketing and digital analytics services. His complete bio and CV are available at www.ashpahwa.com. Dr. Pahwa earned his doctorate in Computer Science from the Illinois Institute of Technology in Chicago. He is listed in Who's Who in the Frontiers of Science and Technology. He is also a Google Certified Analytics Consultant. His expertise includes search engine optimization, digital analytics, web programming, digital image processing, database management, digital video, and data storage technologies. In Industry, Dr. Pahwa has worked for General Electric, AT&T Bell Laboratories, Xerox Corporation, and Oracle. He founded CD-Gen, Inc. and DV Studio Technologies, LLC., which introduced successful products for CD-Recording (CDR) and MPEG encoding. His book, CD-Recordable Bible was published in English, Japanese, and German. In Academia, Dr. Pahwa teaches internet technology courses and conducts webinars in the University of California system. Since 2008, he taught many courses at UC Irvine, UCLA, and UC San Diego, including:
2 Website Development Digital Marketing Predictive Analytics WordPress CMS Search Engine Optimization R Programming Microsoft ASP.NET Digital Analytics Predictive Analytics Using Google Analytics Google AdWords MATLAB Programming Course Description: Digital Analytics (DA) is a set of business and technical activities that create and collect Big Data and process it into analyses, recommendations, optimizations, and predictions. Digital Analytics (DA) provides sophisticated traffic information about a web service and it delivers a comprehensive array of business intelligence and visitor behavior insights which is vital for any marketing department. DA data can be used for market predictions. The goal of this course is to teach how to effectively use DA data by building predictive models. The course will focus on the marketing applications of Predictive Analytics (PA). One of the difficult tasks in PA is the acquisition of high-quality input data. If the input data is inaccurate or incomplete, the predictive modeling results will be equally flawed. Since analytics data is comprehensive, this resource can be used very effectively to predict certain future events or trends. Microsoft Excel and R scripting language will be used as a work engine for data analysis. PA is a leading-edge technology that is being adopted by many successful Fortune 500 corporations. As the name suggests, it seeks to predict the outcome of certain events. Predictive analytics is derived from machine learning and data mining techniques. PA is based on classical statistical techniques such as Decision trees, Linear and logistic regression, Naïve Bayes statistics, and Clustering. All of these techniques will be discussed in this course using marketing applications. Twitter data will be analyzed to compute the customer sentiments. Recommender systems used by most online retailers will be discussed. This course is designed for marketing professionals who are currently working with Digital Analytics. It will provide insight into how analytics data can be used for market predictions. Course Objectives: The goal of this course is to examine the predictive power of analytics data for marketing applications. Students will learn the following: Predictive Analytics for marketing applications. Analytics as the source of data for Predictive Analytics. Predictive Analytics techniques. Excel and R statistical packages for data analysis. Problems marketing professionals face using Predictive Analytics. Expected Learning Outcomes: Data preparation techniques for Predictive Analytics projects Identifying the Predictive Analytics technique used for a specific project Understanding Predictive Analytics techniques Identifying marketing application for predictive analytics Accessing different Predictive Analytics techniques Materials: Class notes will be provided in PDF format
3 Required Texts and Materials: There is no text book for this course. Class material will be taken from applicable white papers, books and websites. The source of material will be disclosed to the students before every class. Optional Texts and Materials: None Grading: Course grades will be based on participation and completion of assignments as follows: Evaluation of Student Performance: Homework assignments Discussion Forum Questions Total 100 points 50 Points 150 points Grading Scale: 90 100 = A 80 89 = B 70 79 = C 60 69 = D Below 60 = F Please note that ALL COURSE GRADES ARE FINAL. Deliverables/Assignments: Each assignment that students need to submit should have its own description, delineating what the assignment entails.
4 Expectations: Students are expected to: Actively participate in class discussions Complete all readings and homework as assigned Be on time in submitting homework assignments Communicate respectfully to instructors and fellow classmates Utilize professional level English in presentations and written assignments Policies: Incompletes: The interim grade Incomplete may be assigned when a student's work is of passing quality, but a small portion of the course requirements is incomplete for good cause (e.g. illness or other serious problem). It is the student s responsibility to discuss with the instructor the possibility of receiving an I grade as opposed to a non passing grade. The student is entitled to replace this grade by a passing grade and to receive unit credit provided they complete the remaining coursework satisfactorily, under the supervision of and in a time frame determined by the instructor in charge, but in no case later than the end of the next academic quarter. At that time, the Registrar will cause all remaining Incompletes to lapse to the grade "F". Note: Receiving an I does not entitle a student to retake all or any part of the course at a later date. Academic Honesty Policy Academic dishonesty covers behavior in cheating, plagiarism, and fabrication of information. These behaviors are not tolerated. Students are encouraged to familiarize themselves with the UCLA Extension Student Conduct Code and the official statements regarding cheating and plagiarism at: https://www.uclaextension.edu/pages/str/studentconduct.aspx Services for Students with Disabilities In accordance with Section 504 of the Rehabilitation Act of 1973 and the Americans with Disabilities Act of 1990, UCLA Extension provides appropriate accommodations and support services to qualified applicants and students with disabilities. These include, but are not limited to, auxiliary aids/services, such as note takers, audiotaping of courses, sign language interpreters, and assistive listening devices for hearing impaired individuals, extended time for and proctoring of exams, and registration assistance. Accommodations and types of support services vary and are specifically designed to meet the disability related needs of each stude3nt based on current, verifiable medical documentation. Arrangements for auxiliary aids/services are available only through UCLA Extension Disabled Student Services at (310) 825 4581 (voice/tty) or by email at access@uclaextension.edu. Please request such arrangements with at least five working days advance notice. All assistance is handled in confidence. Accommodations must be pre approved. Requests for retroactive accommodation will not be accepted.
5 Course Outline Lesson Date Subject Marketing Application 1 What is Predictive Analytics? CRISP/DM method. Data Problems in Predictive Analytics o Inaccuracy, Reliability, Validity, Missing data, Sparse data, Non-stationary data, Over fitting 2 Data Source Google Analytics o Key Metrics o Key Performance Indicators Twitter Other Marketing Data Sources for PA 3 Tools for Predictive Analytics Excel, R, KNIME R interface to Google Analytics & Twitter Statistical Significance of Results Basic Statistics 4 Linear Regression 2 variables Multi-variable 5 Logistic Regression Sales Forecasting Customer Requirements 2 variables Multi-variable 6 Naïve Bayes Predictive Analytics method Marketing Research 7 Decision Trees Predictive Analytics method Market Segmentation 8 Data Clustering Market Segmentation 9 Sentiment Analysis of Twitter data using Predictive Analytics Customer Sentiment Analysis 10 Recommender Systems Sales and Marketing 11 Model Assessment: Comparison of modeling techniques