Eray Turkel

Eray Turkel

Senior Data Scientist, Google

Publications

You can also see my publications on my Google Scholar profile.

Publications

Eray Turkel, Anish Saha, Rhett C. Owen, Greg J. Martin, and Shoshana Vasserman

(Press Coverage- Non technical) (Dataset and Code)

We develop a language model to measure the investigative content of news articles, using a large database of newspaper articles between 2010-2018. Our method combines an unsupervised document influence model with supervised classification based on award winning investigative articles. The model achieves a recall of 67% on unseen test data between 2018-2020.

Eray Turkel

I mathematically characterize the optimal mechanisms in a market environment where a decision maker is designing an auction to allocate ads in an online platform. The decision maker’s preferences take into account both the ad revenues collected by the platforms, and other societal objectives such as ad diversity and price equity. I use Twitter’s political advertising database to empirically estimate bidding and valuation distributions of online advertisers using a Bayesian statistical model and simulate equilibria under different auction mechanisms.

Avidit Acharya, Edoardo Grillo, Takuo Sugaya, and Eray Turkel

We build a game theoretic model of electoral campaigns as dynamic contests in which two candidates allocate their advertising budgets over time to affect their odds of winning, modeled as a mean-reverting stochastic process. We use our characterization of the equilibrium spending path to statistically estimate the over-time rate of decay in the effectiveness of advertising in actual campaigns, using a large dataset of TV advertising from the US between 2000-2014.

Working Papers

Itai Ater, Adi Shany, Brad Ross, Eray Turkel, Shoshana Vasserman (In preperation)

We analyze a major field experiment testing the efficacy of congestion pricing fees in Israel using a panel dataset. Our analysis shows that individuals cut their congestion inducing driving in response to per-Km pricing. We focus on heterogeneity in effects and implications on highway traffic density.

Yunus Can Aybas and Eray Turkel (Revise & Resubmit, last updated Dec 2022)

We analyze the mathematical foundations of games of coarse communication with commitment. Our model captures interactions between a sender and a receiver, where the sender is unable to fully describe the state of the world or recommend all possible actions due to communication constraints, such as credit ratings, product ratings, or advertising. We characterize optimal ways to send information using limited signals, show that the sender’s optimization problem can be tractably solved by a finite program, and prove a general upper bound on the marginal value of a signal.

Posters, Presentations, and Other Work

Mine Su Erturk and Eray Turkel

(MIT Conference on Digital Experimentation, November 2020)

We study a setting where a decision maker is conducting experiments in a network environment, where other analysts are simulataneously conducting experiments on the same network. We analyze an experimenter’s problem where the goal is learning an optimal treatment regime over the network while limiting the contamination on other experimenters. We provide theoretical regret bounds and study the performance of our suggested policy through simulations.

Eray Turkel and Ignacio Martinez

(Online Full Text)

Businesses continuously seek ways to measure the impact of their interventions, such as marketing campaigns, pricing adjustments, or product launches. The difficulty arises in discerning the intervention’s effect from the natural fluctuations of business metrics. Bayesian structural time series emerges as a tool that, when employed judiciously, can leverage data to steer your decisions. However, if used without grasping the underlying assumptions, it can lead to misguided choices while creating a false sense of data-driven decision-making.