|
Data Science and Analytics |
|
Course Description The seminar explores - Leading-edge trends in Data Science and Analytics at Silicon Valley and tech firms
The speakers will include executives, entrepreneurs, and researchers from leading firms.
The topics covered will include (a subset of): - Data Analytics and Big Data - Machine Learning and scalability - Business Analytics including Online Marketing and Advertising, Financial Services and Risk Analytics, Operational and Service Analytics - Information Retrieval (Search) - Information Extraction - Social Networks and Social Media - Healthcare Analytics - Energy Analytics
The seminar will cover the following aspects: - The types of problems being addressed in data science and analytics, the component methods and technologies being developed, and fruitful areas for research and entrepreneurial efforts - This requires attendance and participation in the seminar series and is open to the broader student and faculty community
Prerequisites: None
Units: 2
Course Objectives 1. Learn about and understand landscape of Data Science and Analytics including a subset of - The impact of Data Analytics on Business Analytics - Data Analytics including Machine Learning, Data Mining, Text/Image/Video Mining and Analytics, Search - Information Retrieval (IR), Social Networks, Web Analytics, Online Advertising and Computational Marketing - Implications of Big Data for Analytics
2. Develop - A perspective on the business needs, state of the art research and technology, gaps, and emerging and novel mathematical and other techniques and approaches to address these - A report summarizing this perspective, including research approaches, product/service or entrepreneurial opportunities in some detail - Detail for research would include problem description, model formulation, and developed solution approaches, with validation on real data, where feasible - Detail for product/service or entrepreneurial opportunities would include the details of a business case, business model, competitive analysis, positioning, market analysis (and study where feasible etc.) - Weekly bullet point updates, monthly 5 page updates, and a final 20 (15-25) page report
Course Approach to accommodate heterogeneous class Students can function in one of three modes: - Prior MS/PhD level training in Machine Learning (ML)and/or Data Mining (DM) or math/econ/IEOR analytics training - Students with no prior ML/DM background, concurrently studying i290: Data Mining and Analytics (several iSchool students+ others are in this category) - Students with no prior background, and not concurrently studying ML/DM: These students can be provided some Black Box model/software usage training by Jimi Shanahan, to better appreciate the seminar talks, and to be able to work on the report.
Alternate Backgrounds, Goals for Seminar, and Fit 1. General training and exposure, and concurrent i290 2. Deeper machine learning or IEOR background The goals might then either be more business or entrepreneurship oriented, or more research oriented reports/projects for the course.
Class Report/Project 1. Given a business/data analytics problem, decompose overall problem and solution into standard components, including unsupervised (e.g clustering, Principal Components), supervised (e.g. classification, prediction), optimization (business analytics), Reinforcement Learning etc. 2. Identify in a talk, or in reading a mathematical/technical or white paper, the analytics used to solve a given business problem, which use the commodity algorithms, versus those which adapt or combine these commodity algorithms in a novel way, or create novel algorithms. 3. For a given industrial research project, or analytic product/service, or entrepreneurial concept/product/service, identify both an analytic approach or methodology gap, and a creative way in which the gap can be met. Also, identify the business model which will result in financial success.
Grades 10% class participation, very short bullet point highlights of talks, and interaction; 90% on weekly (15%), monthly (25%), and final submission of reports (40%), and PPT presentations (10%). Crisp, insightful, analytic reports will receive higher points.
|
|
ISchool 296A Spring 2012 |
|
Home Lecture Time: Wednesdays 2-4 PM Location: 205 South Hall
Instructors and Contact Information Ram Akella, Ray Larson, Industrial Expert: James Shanahan Email: akella@ischool.berkeley.edu, Telephone: 650-279-3078 Skype: ramakella1 Office: 422 Sutardja Dai Hall Office Hours: By appointment in the 4-6 pm window (or variants) on Wednesday (or by phone). |
|
|