Filters








8 Hits in 0.43 sec

Omnipredictors [article]

Parikshit Gopalan, Adam Tauman Kalai, Omer Reingold, Vatsal Sharan, Udi Wieder
2021 arXiv   pre-print
We introduce the notion of an (ℒ,𝒞)-omnipredictor, which could be used to optimize any loss in a family ℒ.  ...  In a sense, omnipredictors extract all the predictive power from the class 𝒞, irrespective of the loss function in ℒ.  ...  Hence f S is an (L, C)-omnipredictor.  ... 
arXiv:2109.05389v1 fatcat:4chefbjh5vfl5db64xyht6ut5e

Cost effective speculation with the omnipredictor

Arthur Perais, André Seznec
2018 Proceedings of the 27th International Conference on Parallel Architectures and Compilation Techniques - PACT '18  
All together: the Omnipredictor The TAGE predictor provides a prediction for each instruction.  ...  Then, we unify this indirect target prediction scheme with MDP and conditional prediction within the omnipredictor infrastructure.  ... 
doi:10.1145/3243176.3243208 dblp:conf/IEEEpact/PeraisS18 fatcat:py3lqseac5cs7jnlwcjdrlpbzm

Omnipredictors

Parikshit Gopalan, Adam Tauman Kalai, Omer Reingold, Vatsal Sharan, Udi Wieder, Mark Braverman
2022
The predictors are even omnipredictors with respect t [...]  ...  We introduce the notion of an (L,𝒞)-omnipredictor, which could be used to optimize any loss in a family L.  ...  Hence f S is an (L, C)-omnipredictor.  ... 
doi:10.4230/lipics.itcs.2022.79 fatcat:snluopsfvjbhpbaceflafoakvq

Front Matter, Table of Contents, Preface, Conference Organization

Mark Braverman
2022
. . . . . 77:1-77:19 Testing Distributions of Huge Objects Oded Goldreich and Dana Ron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78:1-78:19 Omnipredictors  ... 
doi:10.4230/lipics.itcs.2022.0 fatcat:fx5ntytpezfgffrwk6o6k5tila

Leveraging Value Equality Prediction for Value Speculation

Kleovoulos Kalaitzidis, André Seznec
2020 ACM Transactions on Architecture and Code Optimization (TACO)  
Finally, one could also consider the ETAGE equality predictor combined with the Omnipredictor [19] , which may predict branches and memory dependencies within the same predictor.  ... 
doi:10.1145/3436821 fatcat:me4ailflljdpridpjstqef5xv4

Decision-Making under Miscalibration [article]

Guy N. Rothblum, Gal Yona
2022 arXiv   pre-print
., 2021] proposed the notion of an omnipredictor, which takes this idea one step further and seeks a single predictor that can be post-processed to be competitive with a large collection of loss functions  ... 
arXiv:2203.09852v1 fatcat:y7oj4gvwxncw5cwutekbvrpyim

Low-Degree Multicalibration [article]

Parikshit Gopalan, Michael P. Kim, Mihir Singhal, Shengjia Zhao
2022 arXiv   pre-print
E (f (x) − f * (x)) 2 ≤ min c∈C E (c(x) − f * (x)) 2 This loss minimization guarantee is similar to the "omnipredictor" loss minimization guarantees from full multicalibration, recently established in  ... 
arXiv:2203.01255v2 fatcat:rcm7mpbhlrgiriplffxxm3yiwa

Beyond Bernoulli: Generating Random Outcomes that cannot be Distinguished from Nature

Cynthia Dwork, Michael P. Kim, Omer Reingold, Guy N. Rothblum, Gal Yona
2022 International Conference on Algorithmic Learning Theory  
They show that any multicalibrated predictor is also an omnipredictor, formally establishing a sense in which multicalibration can be viewed as a strengthening of the standard agnostic PAC solution concept  ... 
dblp:conf/alt/DworkKRRY22 fatcat:zqg2ga3v2bdxvpl7itwkmjq6na