Joel Barajas

Ph.D Candidate
Department of Electrical Engineering
University of California Santa Cruz

Email: jbarajas at soe dot ucsc dot edu

LinkedIn user link, Google Scholar Citations link

News

• 05/2015: I have presented the poster titled Evaluating User Targeting Policies: Simulation Based on Randomized Experiment Data, at the 24th International World Wide Web Conference, WWW 2015. Here is the poster. Questions are welcome!!
• 05/2015: I have presented the paper Estimating Ad Impact on Clicker Conversions for Causal Attribution: A Potential Outcomes Approach, at the SIAM International Conference on Data Mining 2015, SDM 15 (program). Here are the DOI, and the presentation slides. Questions are welcome!!
• 03/2015: I have received the SIAM Student Travel Award to attend the SIAM International Conference on Data Mining (SDM15), to be held April 30 – May 2, 2015, in Vancouver, BC, Canada.
• 02/2015: We have a poster accepted at the WWW 2015: Evaluating User Targeting Policies: Simulation Based on Randomized Experiment Data. pdf, Link.
• 12/2014: We have a paper accepted for presentation at the SIAM Internal Conference in Data Mining (SDM) 2015: Estimating Ad Impact on Clicker Conversions for Causal Attribution: A Potential Outcomes Approach. pdf, Link.

Brief Professional Description

Primary student of a collaborative project between UC Santa Cruz and AOL Advertising.com to evaluate display advertising and make this evaluation more feasible technically and economically since 2010. Here a related AOL Official Blog post.

Going from Data Mining and Bayesian Statistics background, to design of experiments, causality and evaluation in general. Trying to figure out the interception in the context of Online Advertising. See Publications.

3 years of Teaching assistant experience (TA) at UC Santa Cruz in Data Mining, Web Mining, Optimization, Computational Advertising (taught by yahoo! scientists Andrei Broder and Vanja Josifovski link). Regular student audience is composed of Grad students and industry engineers from the Silicon Valley.

Statistical Topic Modeling work to model topic correlations with feasible variants of LDA (with conjugate priors) jointly with Karla L Caballero and Ram Akella.

Summer intern at AT&T interactive (yellow pages) modeling user reviews with statistical topic modeling in order to improve local search under the supervision of Jimi Shanahan.