Optimization Online


Coalescing Data and Decision Sciences for Analytics

Yunxiao Deng(yunxiaod***at***usc.edu)
Junyi Liu(junyiliu***at***usc.edu)
Suvrajeet Sen(s.sen***at***usc.edu)

Abstract: The dream of analytics is to work from common, clean, and consistent data sources in a manner that all of its facets (descriptive, predictive, and prescriptive) are sup- ported via a coherent vision of data and decision sciences. To the extent that data and decisions sciences work within logically/mathematically consistent frameworks, and that these paradigms operate within computationally realistic structures, ana- lytics will help organizations thrive. Done right, analytics promises to plant more OR/MS resources at C-suites where major corporate decisions are made. Done poorly, the OR/MS community will look back at these days as a lost opportunity to make a long-term difference to both theory and practice of analytics. With this tutorial, we invite the OR/MS community to join an intellectually vibrant endeavor to inte- grate seemingly disparate pillars, namely predictive and prescriptive analytics. Just as importantly, we show how OR/MS strengths in modeling, algorithms, and appli- cations can facilitate coalescing the data and decision sciences. We draw upon several examples which pair data science tools such as filtering and regression, to decision sci- ence tools such as Monte Carlo tree search and stochastic programming. The marriage of these two paradigms (data and decision sciences) is what we refer to as Learn- ing Enabled Optimization (LEO). This tutorial covers fundamental concepts for the fusion of statistical and optimization modeling, statistically approximate optimality, sampling-based algorithms, and finally, model assessment and selection.

Keywords: Statistical Learning, Optimization, Model Assessment and Selection

Category 1: Stochastic Programming


Download: [PDF]

Entry Submitted: 05/18/2018
Entry Accepted: 05/19/2018
Entry Last Modified: 05/18/2018

Modify/Update this entry

  Visitors Authors More about us Links
  Subscribe, Unsubscribe
Digest Archive
Search, Browse the Repository


Coordinator's Board
Classification Scheme
Give us feedback
Optimization Journals, Sites, Societies
Mathematical Optimization Society