The Economics of Artificial Intelligence

Home > Other > The Economics of Artificial Intelligence > Page 92
The Economics of Artificial Intelligence Page 92

by Ajay Agrawal


  gate .net/ publication/ 321902459_Behavior_Revealed_in_Mobile_Phone_Usage

  _Predicts_Loan_Repayment.

  Blei, D. M. 2012. “Probabilistic Topic Models.” Communications of the ACM 55

  (4): 77.

  Blei, D. M., A. Y. Ng, and M. I. Jordan. 2003. “Latent Dirichlet Allocation.” Journal of Machine Learning Research, 3 (Jan): 993– 1022.

  Chapelle, O., and L. Li. 2011. “An Empirical Evaluation of Thompson Sampling.”

  Proceedings of the Conference on Neural Information Processing Systems. https://

  papers.nips.cc/ paper/ 4321-an- empirical- evaluation- of-thompson- sampling.

  Chernozhukov, V., D. Chetverikov, M. Demirer, E. Duo, C. Hansen, and W. Newey.

  2017. “Double/ Debiased/ Neyman Machine Learning of Treatment Eff ects. Janu-

  ary. Cornell University Library. http:// arxiv .org/ abs/ 1701.08687.

  Chernozhukov, V., C. Hansen, and M. Spindler. 2015. “Valid Post- Selection and

  Post- Regularization Inference: An Elementary, General Approach.” January.

  www .doi .org/ 10.1146/ annurev- economics- 012315-015826.

  Dimakopoulou, M., S. Athey, and G. Imbens. 2017. “Estimation Considerations in

  Contextual Bandits.” Cornell University Library. https:// arxiv .org/ abs/ 1711.07077.

  Doudchenko, N., and G. W. Imbens. 2016. “Balancing, Regression, Diff erence-

  in-Diff erences and Synthetic Control Methods: A Synthesis.” Technical report,

  National Bureau of Economic Research.

  The Impact of Machine Learning on Economics 545

  Dudik, M., D. Erhan, J. Langford, and L. Li. 2014. “Doubly Robust Policy Evalu-

  ation and Optimization.” Statistical Science 29 (4): 485– 511.

  Dudik, M., J. Langford, and L. Li. 2011. “Doubly Robust Policy Evaluation and

  Learning.” International Conference on Machine Learning.

  Dufl o, E. 2017. “The Economist as Plumber.” NBER Working Paper no. 23213,

  Cambridge, MA.

  Eckles, D., B. Karrer, J. Ugander, L. Adamic, I. Dhillon, Y. Koren, R. Ghani,

  P. Senator, J. Bradley, and R. Parekh. 2016. “Design and Analysis of Experiments

  in Networks: Reducing Bias from Interference.” Journal of Causal Inference 1– 62.

  www .doi .org/ 10.1515/ jci- 2015-0021.

  Egami, N., C. Fong, J. Grimmers, M. Roberts, and B. Stewart. 2016. “How to Make

  Causal Inferences Using Text.” Working paper. https:// polmeth.polisci.wisc .edu

  / Papers/ ais .pdf.

  Engstrom, R., J. Hersh, and D. Newhouse. 2017. “Poverty from Space: Using High-

  Resolution Satellite Imagery for Estimating Economic Well- Being (English).”

  Policy Research Working Paper no. WPS 8284, Washington, DC, World Bank

  Group.

  Feraud, R., R. Allesiardo, T. Urvoy, and F. Clerot. 2016. “Random Forest for the Con-

  textual Bandit Problem.” Proceedings of Machine Learning Research 51:93– 101.

  Glaeser, E. L., A. Hillis, S. D. Kominers, and M. Luca. 2016. “Predictive Cities

  Crowdsourcing City Government: Using Tournaments to Improve Inspection

  Accuracy.” American Economic Review 106 (5): 114– 18.

  Glaeser, E. L., S. D. Kominers, M. Luca, and N. Naik. 2015. “Big Data and Big

  Cities: The Promises and Limitations of Improved Measures of Urban Life.”

  NBER Working Paper no. 21778, Cambridge, MA.

  Glaeser, E. L., S. D. Kominers, M. Luca, and N. Naik. 2018. “Big Data and Big

  Cities: The Promises and Limitations of Improved Measures of Urban Life.”

  Economic Inquiry 56 (1): 114– 37.

  Goel, S., J. M. Rao, and R. Shroff . 2016. “Precinct or Prejudice? Understanding

  Racial Disparities in New York City’s Stop- and- Frisk Policy.” Annals of Applied

  Statistics 10 (1): 365– 94.

  Goldenshluger, A., and A. Zeevi. 2013. “A Linear Response Bandit Problem.” Sto-

  chastic Systems 3 (1): 230– 61.

  Goldman, M., and J. M. Rao. 2014. “Experiments as Instruments: Heteroge-

  neous Position Eff ects in Sponsored Search Auctions.” The Third Conference

  on Auctions, Market Mechanisms and Their Applications. www .doi .org/ 10.4108

  / eai.8-8-2015.2261043.

  Gopalan, P., J. M. Hofman, and D. M. Blei. 2015. “Scalable Recommendation with

  Hierarchical Poisson Factorization.” Proceedings of the Thirty- First Conference

  on Uncertainty in Artifi cial Intelligence, 326– 35.

  Hahn, J. 1998. “On the Role of the Propensity Score in Effi

  cient Semiparametric

  Estimation of Average Treatment Eff ects.” Econometrica:315– 31.

  Hartford, J., G. Lewis, and M. Taddy. 2016. “Counterfactual Prediction with

  Deep Instrumental Variables Networks.” Working paper. https:// arxiv .org/ pdf

  / 1612.09596 .pdf.

  Imai, K., and M. Ratkovic. 2013. “Estimating Treatment Eff ect Heterogeneity in

  Randomized Program Evaluation.” Annals of Applied Statistics 7 (1): 443– 70.

  Imbens, G. W., and D. B. Rubin. 2015. Causal Inference in Statistics, Social, and

  Biomedical Sciences. Cambridge: Cambridge University Press.

  Imbens, G. W., and J. M. Wooldridge. 2009. “Recent Developments in the Econo-

  metrics of Program Evaluation.” Journal of Economic Literature 47 (1): 5– 86.

  546 Susan Athey

  Jean, N., M. Burke, M. Xie, W. M. Davis, D. B. Lobell, and S. Ermon. 2016. “Com-

  bining Satellite Imagery and Machine Learning to Predict Poverty.” Science 353

  (6301): 790– 94.

  Jiang, N., and L. Li. 2016. “Doubly Robust Off - Policy Value Evaluation for Rein-

  forcement Learning.” Proceedings of the 33rd International Conference on Machine

  Learning, vol. 48, 652– 61.

  Kallus, N. 2017. “Balanced Policy Evaluation and Learning.” Cornell University

  Library. https:// arxiv .org/ abs/ 1705.07384.

  Kitagawa, T., and A. Tetenov. 2015. “Who Should Be Treated? Empirical Welfare

  Maximization Methods for Treatment Choice.” Technical report, Centre for

  Microdata Methods and Practice, Institute for Fiscal Studies.

  Kleinberg, J., J. Ludwig, S. Mullainathan, and Z. Obermeyer. 2015. “Prediction

  Policy Problems.” American Economic Review 105 (5): 491– 95.

  Kleinberg, J., S. Mullainathan, and M. Raghavan. 2016. “Inherent Trade- Off s in

  the Fair Determination of Risk Scores.” Cornell University Library. https:// arxiv

  .org/ abs/ 1609.05807.

  Komarova, T., D. Nekipelov, and E. Yakovlev. 2015. “Estimation of Treatment

  Eff ects from Combined Data: Identifi cation versus Data Security.” In Economic

  Analysis of the Digital Economy, edited by A. Goldfarb, S. M. Greenstein, and

  C. Tucker, 279– 308. Chicago: University of Chicago Press.

  Künzel, S., J. Sekhon, P. Bickel, and B. Yu. 2017. “Meta- Learners for Estimating

  Heterogeneous Treatment Eff ects Using Machine Learning.” Cornell University

  Library. https:// arxiv .org/ abs/ 1706.03461.

  Laff ont, J.-J., H. Ossard, and Q. Vuong. 1995. “Econometrics of First- Price Auc-

  tions.” Econometrica: Journal of the Econometric Society 63 (4): 953– 80.

  Lewis, R. A., and J. M. Rao. 2015. “The Unfavorable Economics of Measuring the

  Returns to Advertising.” Quarterly Journal of Economics 130 (4): 1941– 73.

  Li, L., S. Chen, J. Kleban, and A. Gupta. 2014. “Counterfactual Estimation and

  Optimization of Click Metrics for Search Engines.” Cornell University Library.

  https:// arxiv .org/ abs/ 1403.1891.

  Li, L., W. Chu, J.
Langford, T. Moon, and X. Wang. 2012. “An Unbiased Offl

  ine

  Evaluation of Contextual Bandit Algorithms with Generalized Linear Models.”

  Journal of Machine Learning Research Workshop and Conference Proceedings

  26:19– 36.

  Li, L., W. Chu, J. Langford, and R. Schapire. 2010. “A Contextual- bandit Approach

  to Personalized News Article Recommendation.” International World Wide Web

  Conference. https:// dl.acm .org/ citation .cfm?doid=1772690.1772758.

  Li, L., Y. Lu, and D. Zhou. 2017. “Provably Optimal Algorithms for Generalized

  Linear Contextual Bandits.” International Conference on Machine Learning.

  https:// arxiv .org/ abs/ 1703.00048.

  McFadden, D. 1973. “Conditional Logit Analysis of Qualitative Choice Behavior.”

  In Frontiers in Econometrics, edited by P. Zarembka. New York: Wiley.

  Mullainathan, S., and J. Spiess. 2017. “Machine Learning: An Applied Econometric

  Approach.” Journal of Economic Perspectives 31 (2): 87– 106.

  Naik, N., S. D. Kominers, R. Raskar, E. L. Glaeser, and C. A. Hidalgo. 2017. “Com-

  puter Vision Uncovers Predictors of Physical Urban Change.” Proceedings of the

  National Academy of Sciences 114 (29): 7571– 76.

  Naik, N., J. Philipoom, R. Raskar, and C. Hidalgo. 2014. “Streetscore- Predicting the

  Perceived Safety of One Million Streetscapes.” In Proceedings of the IEEE Con-

  ference on Computer Vision and Pattern Recognition Workshops, 779– 85.

  Ng, A. Y., and S. J. Russell. 2000. “Algorithms for Inverse Reinforcement Learning.”

  The Impact of Machine Learning on Economics 547

  In Proceedings of the Seventeenth International Conference on Machine Learning, 663– 70. https:// dl.acm .org/ citation .cfm?id=657801.

  Peysakhovich, A., and A. Lada. 2016. “Combining Observational and Experimen-

  tal Data to Find Heterogeneous Treatment Eff ects.” Cornell University Library.

  http:// arxiv .org/ abs/ 1611.02385.

  Robinson, P. M. 1988. “Root- n- Consistent Semiparametric Regression.” Economet-

  rica: Journal of the Econometric Society 56 (4): 931– 54.

  Roth, A. E. 2002. “The Economist as Engineer: Game Theory, Experimentation,

  and Computation as Tools for Design Economics.” Econometrica 70 (4): 1341– 78.

  Ruiz, F. J., S. Athey, and D. M. Blei. 2017. “Shopper: A Probabilistic Model of Con-

  sumer Choice with Substitutes and Complements.” Cornell University Library.

  https:// arxiv .org/ abs/ 1711.03560.

  Scott, S. L. 2010. “A Modern Bayesian Look at the Multi- Armed Bandit.” Applied

  Stochastic Models in Business and Industry 26 (6): 639– 58.

  Strehl, A., J. Langford, L. Li, and S. Kakade. 2010. “Learning from Logged Implicit

  Exploration Data.” Proceedings of the 23rd International Conference on Neural

  Information Processing Systems, vol. 2, 2217– 25.

  Swaminathan, A., and T. Joachims. 2015. “Batch Learning from Logged Bandit

  Feedback through Counterfactual Risk Minimization.” Journal of Machine

  Learning Research 16 (Sep.): 1731−55.

  Thomas, P., and E. Brunskill. 2016. “Data- Effi

  cient Off - Policy Policy Evaluation

  for Reinforcement Learning.” Proceedings of the 33rd International Conference

  on Machine Learning, vol. 48, 2139– 48.

  Tibshirani, R., and T. Hastie. 1987. “Local Likelihood Estimation.” Journal of the

  American Statistical Association 82 (398): 559– 67.

  Ugander, J., B. Karrer, L. Backstrom, and J. Kleinberg. 2013. “Graph Cluster Ran-

  domization.” In Proceedings of the 19th ACM SIGKDD International Conference

  on Knowledge Discovery and Data Mining—KDD ’13, 329. New York: ACM Press.

  ISBN 9781450321747. doi: 10.1145/ 2487575.2487695. http:// dl.acm .org/ citation

  .cfm?doid=2487575.2487695

  van der Laan, M. J., and D. Rubin. 2006. “Targeted Maximum Likelihood Learn-

  ing.” Working Paper no. 213, UC Berkeley Division of Biostatistics.

  Varian, H. R. 2014. “Big Data: New Tricks for Econometrics.” Journal of Economic

  Perspectives 28 (2): 3– 27.

  Wager, S., and S. Athey. Forthcoming. “Estimation and Inference of Heterogeneous

  Treatment Eff ects Using Random Forests.” Journal of the American Statistical

  Association.

  Wan, M., D. Wang, M. Goldman, M. Taddy, J. Rao, J. Liu, D. Lymberopoulos, and

  J. McAuley. 2017. “Modeling Consumer Preferences and Price Sensitivities from

  Large- Scale Grocery Shopping Transaction Logs.” In Proceedings of the 26th

  International Conference on World Wide Web, 1103– 12.

  White, H. 1992. Artifi cial Neural Networks: Approximation and Learning Theory.

  Hoboken, NJ: Blackwell Publishers.

  Yeomans, M., A. K. Shah, and J. Kleinberg. 2016. “Making Sense of Recommenda-

  tions.” Working paper, Department of Economics, Harvard University. https://

  scholar.harvard .edu/ fi les/ sendhil/ fi les/ recommenders55 .pdf.

  Zeileis, A., T. Hothorn, and K. Hornik. 2008. “Model- Based Recursive Partition-

  ing.” Journal of Computational and Graphical Statistics 17 (2): 492– 514.

  Zubizarreta, J. R. 2015. “Stable Weights That Balance Covariates for Estimation

  with Incomplete Outcome Data.” Journal of the American Statistical Association

  110 (511): 910– 22.

  548 Mara Lederman

  Comment Mara Lederman

  Athey provides a comprehensive, accessible, and exciting summary of the

  impact that machine learning (ML) is having—and will continue to have—

  on the fi eld of economics. It is a thorough, thoughtful, and optimistic chap-

  ter that makes clear the unique strengths of ML and the unique strengths

  of traditional econometrics- based techniques for causal inference and high-

  lights both the opportunities to combine these approaches as well as the sorts

  of tasks and problems that are likely to remain in each domain. The chapter

  contains several useful and practical examples that illustrate the application

  of ML techniques to questions and problems that are of interest to econo-

  mists including allocating health care procedures, pricing, and measuring

  the impact of advertising.

  At a broad level, the chapter has four main sections. The chapter begins

  by off ering straightforward defi nitions of unsupervised and supervised ML.

  Athey puts it quite simply: unsupervised ML uses algorithms to identify

  observations that are similar in their covariates, while supervised ML uses

  algorithms to predict an outcome variable from observations on covariates.

  It is important to emphasize, and I will return to this, that the observations

  and variables that ML algorithms can handle often do not look like the

  typical quantitative data that economists use in empirical analysis. Both

  unsupervised and supervised machine- learning techniques can be applied

  to text, images, and video. For example, unsupervised ML algorithms can be

  used to identify similar videos (without needing to specify in advance what

  makes these videos similar) or similar restaurant reviews (again, without

  needed to specify which reviews are positive or negative or what words or

  phrases makes a review positive or negative). Supervised ML algorithms

  can be used to predict variables such as the sentiment of a tweet or the slant

  of a newspaper article, without having to specify e
x ante what the relevant

  covariates are.

  The chapter then discusses a number of ways in which off - the- shelf ML

  techniques can be directly integrated into traditional economics research.

  For example, both unsupervised and supervised ML can be used to create

  variables that can be used in standard econometric analyses. In addition,

  ML techniques can be directly applied to what Kleinberg et al. (2015) call

  “prediction policy problems.” These are policy problems or decisions that

  inherently involve a prediction component and, in these cases, ML tech-

  niques may be superior to other statistical methodologies. These problems

  may involve novel sources of so-called “big data”—such as satellite image

  data used in Glaeser et al. (2018)—but need not. They are simply policy

  Mara Lederman is associate professor of strategic management at Rotman School of Man-

  agement, University of Toronto.

  For acknowledgments, sources of research support, and disclosure of the author’s material fi nancial relationships, if any, please see http:// www .nber .org/ chapters/ c14036.ack.

  Comment 549

  problems in which the predicted value of an unknown variable acts an input

  into a decision.

  The third and most substantial section of the chapter discusses the grow-

  ing literature at the intersection of machine learning, statistics, and econo-

  metrics. As Athey puts it, this literature is developing novel methodolo-

  gies that “harass the strengths of ML algorithms to solve causal inference

  problems.” Athey provides details on a number of recent contributions

  in this area, highlighting the parts of the estimation approaches that are

  improved by ML and the parts that continue to rely on traditional econo-

  metric approaches and assumptions. Athey predicts that these techniques

  will soon become commonly used in applied empirical work in economics.

  Finally, the chapter concludes with a discussion of some of the broader

  eff ects that ML might have on the economics profession, beyond the impact

  on the way we do empirical research, including the types of questions econo-

  mists will ask, the degree of cross- disciplinary collaboration, the production

  function for research and the emergence of the “economist as an engineer,”

  working with business and government to implement policies, and experi-

  ments in a digital environment.

 

‹ Prev