Adwords with Unknown Budgets

listen on castbox.fmlisten on google podcastslisten on player.fmlisten on pocketcastslisten on podcast addictlisten on tuninlisten on Amazon Musiclisten on Stitcher



AdWords with Unknown Budgets

In today’s episode, we are joined by Rajan Udwani, an Assistant Professor at the University of California Berkeley. He joins us to discuss his research titled, Adwords with Unknown Budgets.

Rajan began by discussing how the tools for operations research vary based on the optimization problem. He then delved into the optimization problem for AdWords. He discussed how important optimization is during ad allocation to maximize profit. Rajan explained the approaches to modelling the problem of ad allocation. The first he discussed is the forecasting approach while the second is the adversarial approach.

Rajan also spoke about the greedy approach and the bid shading approach. He explained how their algorithms work and the challenges with them. He then discussed how to find an optimized alternative by randomizing the already existing algorithms. 

Rajan explained the model evaluation process for the problem. He corroborated his results with practical applications of his model for AdWords with an unknown budget. Also, he discussed other applications of his model such as in the allocation of online resources. 

Concluding, Rajan discussed two other ideas (throttling and bid scaling) that can better optimize ad allocation. You can follow Rajan on his Google Scholar page or his webpage.

Rajan Udwani is an Assistant Professor of Industrial Engineering and Operations Research at UC Berkeley. He holds a B.Tech in Electrical Engineering from IIT Bombay and a PhD in Operations Research from MIT. He works on algorithms for optimization under uncertainty with a focus on revenue management and pricing. His work has been recognized by junior faculty and student paper awards from the Institute for Operations Research and the Management Sciences (INFORMS).