Andrés Muñoz MedinaAddress: 251 Mercer Street,
7th floor, Room 718
New York, NY, 10012
Email: amunoz88 at gmail dot com
I just finished implementing a QCQP solver for non-convex problems based on DC programming. You can get the R code here .
I am a co-organizer for the 2014 NIPS workshop on Transfer and Multitask Learning: Theory Meets Practice. In addition I have served as the President for the Courant Student Organization and have been the Recruiting Chair at CSplash, an annual event promoting research in mathematics to high-school students. I am also the organizer of the Courant Machine Learning Seminar and for the past three years, I have been a Research Summer Intern at Google where I have worked on the design of algorithms for learning in large-scale scenarios.
I am currently involved in several exciting research projects. My main areas of interest are: learning in auctions (including all connections between Machine Learning and Game Theory), adaptation (including drifting distributions, theory and algorithms for domain adaptation in classification and regression), analysis and extensions of bandit problems (such as bandits with unbounded losses and contextual bandits). In addition, I am interested in all aspects of Machine Learning and Probability theory.
I am a 5th year PhD student at the Courant Institute of Mathematical sciences, working under the advice of Professor Mehryar Mohri . I am interested in Machine Learning, Statistics, Algorithms and Probability. My current research is focused on learning theory applied to market algorithms, domain adaptation and regret minimization algorithms. However I am happy to tackle any problem related to Machine Learning.
I am the organizer of the Courant Machine Learning Seminar and was president of the Courant Student Organization which is in charge of promoting Courant PhD students to industry as well as community outreach events such as CSplash . I really enjoy teaching and my students agree that I can be tough but am always willing to help them.
When I am not doing research or teaching I enjoy biking, running, swimming and traveling.
Mehryar Mohri and Andrés Muñoz Medina
Learning Algorithms for Second-Price Auctions with Reserve . JMLR (submitted) 2014.
- Mehryar Mohri and Andrés Muñoz Medina Learning to select reserve prices in GSP auctions. Submitted to UAI 2015.
Mehryar Mohri and Andrés Muñoz Medina Optimal Regret Minimization in Posted-Price Auctions with Strategic Buyers
Mehryar Mohri and Andrés Muñoz Medina.
Revenue optimization in AdExchange against Strategic Advertisers.
Workshop on Optimizing Customer Lifetime Value in Online Marketing. ICML 2014.
Mehryar Mohri and Andrés Muñoz Medina. Learning Theory and Algorithms for Revenue Optimization in Second-Price Auctions with Reserve. In Proceedings of ICML 2014.
Corinna Cortes, Mehryar Mohri and Andrés Muñoz Medina Adaptation Algorithm and Theory Based on Generalized Discrepancy
Submitted to KDD 2015.
Mehryar Mohri and Andrés Muñoz Medina. New Analysis and algorithm for learning with drifting distributions. In Proceedings of ALT 2012.
Mehryar Mohri, Corinna Cortes and Andrés Muñoz Medina Discrepancy Minimization: Adaptation for Classification, in preparation, 2014.
Mehryar Mohri, Andrés Muñoz Medina and Scott Yang. Maxent Formulation and Solution to the Contextual Bandit Problem, in preparation, 2014.
Mehryar Mohri, Andrés Muñoz Medina and Scott Yang. Bandits with unbounded losses, in preparation, 2014.
Talks and tutorials
- NIPS 2014 Workshop: Learning Theory and Algorithms for Domain Adaptation
- NIPS 2014 Workshop: Revenue Optimization in Posted-Price Auctions against Strategic Buyers
- ICML workshop: Revenue optimization in AdExchange against strategic advertisers
- Learning Theory and Algorithms for Second-Price Auctions with Reserve
- An introduction to DC programming
- Non convex QCQP solver for R: QCQP_1.0 . Please send me an email if you find any bugs in it.