Why Marshall
Leadership
Dean Geoffrey Garrett
Dean's Cabinet
Boards
Real-World Learning
Human Leadership
Tech Fluency
Global Opportunities
Diversity, Equity and Inclusion
Teaching + Innovation
Experiential Learning Center
Open Expression Statement
Programs
Undergraduate Programs
Admissions
Degrees
BS Business Administration (BUAD)
Business Emphases
BS Accounting (ACCT)
World Bachelor in Business (WBB)
BS Business of Cinematic Arts (BCA)
BS Artificial Intelligence for Business (BUAI)
Undergraduate Minors
Graduate Programs
MBA Programs
Full-Time MBA (FTMBA)
Executive MBA (EMBA)
Part-Time MBA (MBA.PM)
International MBA (IBEAR)
Online MBA (OMBA)
Specialized Masters
MS Business Administration (MSBUSAD)
MS Business Analytics (MSBA)
MS Entrepreneurship + Innovation (MSEI)
MS Finance (MSF)
MS Global Supply Chain Management (MSGSCM)
MS Marketing (MSMKT)
MS Social Entrepreneurship (MSSE)
Master of Business for Veterans (MBV)
Master of Management Studies (MMS)
Accounting Masters
Master of Accounting (MAcc)
Master of Business Taxation (MBT)
Master of Business Taxation for Working Professionals (MBT.WP)
PhD Program
Accounting
Data Sciences + Operations
Finance
Management + Organization
Marketing
Graduate Certificates
GC in Business Analytics
GC in Financial Analysis + Valuation
GC in Management Studies
GC in Marketing
GC in Optimization + Supply Chain Management
GC in Strategy + Management Consulting
GC in Sustainability + Business
GC in Technology Commercialization
GC in Library and Information Management – Online
Executive Education Redirect
Departments
Business Communication (BUCO)
Faculty
Data Sciences and Operations (DSO)
Finance + Business Economics (FBE)
Leventhal School of Accounting (ACCT)
Lloyd Greif Center for Entrepreneurial Studies (BAEP)
Management and Organization (MOR)
Marketing (MKT)
Institutes + Centers
Peter Arkley Institute for Risk Management
Brittingham Social Enterprise Lab
Center for Investment Studies
Initiative on Digital Competition
Randall R. Kendrick Global Supply Chain Institute
Center for Effective Organizations
Lloyd Greif Center for Entrepreneurial Studies
VanEck Digital Assets Initiative
Institute for Outlier Research in Business
Center for Global Innovation
Neely Center for Ethical Leadership and Decision Making
Trojan Network
Recruiting
Undergraduate
Graduate
Career Services
Giving + Support
Alumni Engagement + Resources
Student Organizations
Commencement
Jacob Bien's research focuses on statistical machine learning and in particular the development of novel methods that balance flexibility and interpretability for analyzing complex data. He combines ideas from convex optimization and statistics to develop methods that are of direct use to scientists and others with large datasets. His work has been supported by an NSF CAREER award, a three-year NSF grant on high-dimensional covariance estimation, an NIH R01 grant on methods for multi-view data, and a grant from the Simons Foundation on developing new statistical methodology for oceanography. He serves as an associate editor of Biometrika and the Journal of Computational and Graphical Statistics, and he was previously an associate editor for Biostatistics. Before joining USC, he was an assistant professor at Cornell.
Areas of Expertise
NEWS + EVENTS
"Comparing Trained and Untrained Probabilistic Ensemble Forecasts of COVID-19 Cases and Deaths in the United States"
In recent work published in the International Journal of Forecasting, Professor Jacob Bien and his co-authors describe their efforts to evaluate the performance of different ensemble models to forecast cases and deaths as part of the US Covid-19 Forecast Hub, which was used by the CDC and various state health officials.
RESEARCH + PUBLICATIONS
Stability selection (Meinshausen and Buhlmann, 2010) makes any feature selection method more stable by returning only those features that are consistently selected across many subsamples. We prove (in what is, to our knowledge, the first result of its kind) that for data containing highly correlated proxies for an important latent variable, the lasso typically selects one proxy, yet stability selection with the lasso can fail to select any proxy, leading to worse predictive performance than the lasso alone.We introduce cluster stability selection, which exploits the practitioner's knowledge that highly correlated clusters exist in the data, resulting in better feature rankings than stability selection in this setting. We consider several feature-combination approaches, including taking a weighted average of the features in each important cluster where weights are determined by the frequency with which cluster members are selected, which we show leads to better predictive models than previous proposals.We present generalizations of theoretical guarantees from Meinshausen and Buhlmann (2010) and Shah and Samworth (2012) to show that cluster stability selection retains the same guarantees. In summary, cluster stability selection enjoys the best of both worlds, yielding a sparse selected set that is both stable and has good predictive performance.