Variational Dropout and the Local Reparameterization Trick
- A. Blum, Nika Haghtalab, Ariel D. Procaccia
- Computer ScienceNIPS
- 2015
The Variational dropout method is proposed, a generalization of Gaussian dropout, but with a more flexibly parameterized posterior, often leading to better generalization in stochastic gradient variational Bayes.
Commitment Without Regrets: Online Learning in Stackelberg Security Games
- M. Balcan, A. Blum, Nika Haghtalab, Ariel D. Procaccia
- Computer ScienceACM Conference on Economics and Computation
- 15 June 2015
This work designs no-regret algorithms whose regret (when compared to the best fixed strategy in hindsight) is polynomial in the parameters of the game, and sublinear in the number of times steps.
Efficient Learning of Linear Separators under Bounded Noise
- Pranjal Awasthi, M. Balcan, Nika Haghtalab, Ruth Urner
- Computer ScienceAnnual Conference Computational Learning Theory
- 11 March 2015
This work provides the first evidence that one can indeed design algorithms achieving arbitrarily small excess error in polynomial time under this realistic noise model and thus opens up a new and exciting line of research.
Ignorance is Almost Bliss: Near-Optimal Stochastic Matching With Few Queries
- A. Blum, Nika Haghtalab, Ariel D. Procaccia, Ankit Sharma
- Computer Science, MathematicsACM Conference on Economics and Computation
- 15 July 2014
It is shown on both generated data and on real data from the first 169 match runs of the UNOS nationwide kidney exchange that even a very small number of non-adaptive edge queries per vertex results in large gains in expected successful matches.
Learning Optimal Commitment to Overcome Insecurity
- A. Blum, Nika Haghtalab, Ariel D. Procaccia
- Computer ScienceNIPS
- 8 December 2014
This work designs an algorithm that optimizes the defender's strategy with no prior information, by observing the attacker's responses to randomized deployments of resources and learning his priorities.
Collaborative PAC Learning
- A. Blum, Nika Haghtalab, Ariel D. Procaccia, Mingda Qiao
- Computer ScienceNIPS
- 1 December 2017
A collaborative PAC learning model, in which k players attempt to learn the same underlying concept, with an Omega(ln(k)) overhead lower bound, showing that the results are tight up to a logarithmic factor.
Maximizing Welfare with Incentive-Aware Evaluation Mechanisms
- Nika Haghtalab, Nicole Immorlica, Brendan Lucier, Jack Wang
- Computer ScienceInternational Joint Conference on Artificial…
- 1 July 2020
A classification problem where the inputs are controlled by strategic individuals who can modify their features at a cost is studied, and it is shown that the optimal classifier is an appropriate projection of the quality score.
Oracle-Efficient Online Learning and Auction Design
- Miroslav DudÃk, Nika Haghtalab, Haipeng Luo, R. Schapire, Vasilis Syrgkanis, Jennifer Wortman Vaughan
- Computer ScienceIEEE Annual Symposium on Foundations of Computer…
- 5 November 2016
We consider the design of computationally efficient online learning algorithms in an adversarial setting in which the learner has access to an offline optimization oracle. We present an algorithm…
The disparate equilibria of algorithmic decision making when individuals invest rationally
- Lydia T. Liu, Ashia C. Wilson, Nika Haghtalab, A. Kalai, C. Borgs, J. Chayes
- EconomicsFAT*
- 4 October 2019
It is shown that decoupling achieves optimal outcomes in the realizable case but has discrepant effects that may depend on the initial conditions otherwise and subsidizing the cost of investment is shown to create better equilibria for the disadvantaged group even in the absence of realizability.
k-center Clustering under Perturbation Resilience
- M. Balcan, Nika Haghtalab, Colin White
- Computer ScienceInternational Colloquium on Automata, Languages…
- 14 May 2015
This work provides strong positive results both for the asymmetric and symmetric k-center problems under a natural input stability (promise) condition called α-perturbation resilience and provides algorithms that give strong guarantees simultaneously for stable and non-stable instances.
...
...