Justin Hsu, Jamie Morgenstern, Ryan M. Rogers, Aaron Roth, R. Vohra
Walrasian equilibrium prices have a remarkable property: they allow each buyer to purchase a bundle of goods that she finds the most desirable, while guaranteeing that the induced allocation over all buyers will globally maximize social welfare. However, this clean story has two caveats. * First, the prices may induce indifferences. In fact, the minimal equilibrium prices necessarily induce indifferences. Accordingly, buyers may need to coordinate with one another to arrive at a socially optimal outcome---the prices alone are not sufficient to coordinate the market. * Second, although natural procedures converge to Walrasian equilibrium prices on a fixed population, in practice buyers typically observe prices without participating in a price computation process. These prices cannot be perfect Walrasian equilibrium prices, but instead somehow reflect distributional information about the market. To better understand the performance of Walrasian prices when facing these two problems, we give two results. First, we propose a mild genericity condition on valuations under which the minimal Walrasian equilibrium prices induce allocations which result in low over-demand, no matter how the buyers break ties. In fact, under genericity the over-demand of any good can be bounded by 1, which is the best possible at the minimal prices. We demonstrate our results for unit demand valuations and give an extension to matroid based valuations (MBV), conjectured to be equivalent to gross substitute valuations (GS). Second, we use techniques from learning theory to argue that the over-demand and welfare induced by a price vector converge to their expectations uniformly over the class of all price vectors, with respective sample complexity linear and quadratic in the number of goods in the market. These results make no assumption on the form of the valuation functions. These two results imply that under a mild genericity condition, the exact Walrasian equilibrium prices computed in a market are guaranteed to induce both low over-demand and high welfare when used in a new market where agents are sampled independently from the same distribution, whenever the number of agents is larger than the number of commodities in the market.
{"title":"Do prices coordinate markets?","authors":"Justin Hsu, Jamie Morgenstern, Ryan M. Rogers, Aaron Roth, R. Vohra","doi":"10.1145/2897518.2897559","DOIUrl":"https://doi.org/10.1145/2897518.2897559","url":null,"abstract":"Walrasian equilibrium prices have a remarkable property: they allow each buyer to purchase a bundle of goods that she finds the most desirable, while guaranteeing that the induced allocation over all buyers will globally maximize social welfare. However, this clean story has two caveats. * First, the prices may induce indifferences. In fact, the minimal equilibrium prices necessarily induce indifferences. Accordingly, buyers may need to coordinate with one another to arrive at a socially optimal outcome---the prices alone are not sufficient to coordinate the market. * Second, although natural procedures converge to Walrasian equilibrium prices on a fixed population, in practice buyers typically observe prices without participating in a price computation process. These prices cannot be perfect Walrasian equilibrium prices, but instead somehow reflect distributional information about the market. To better understand the performance of Walrasian prices when facing these two problems, we give two results. First, we propose a mild genericity condition on valuations under which the minimal Walrasian equilibrium prices induce allocations which result in low over-demand, no matter how the buyers break ties. In fact, under genericity the over-demand of any good can be bounded by 1, which is the best possible at the minimal prices. We demonstrate our results for unit demand valuations and give an extension to matroid based valuations (MBV), conjectured to be equivalent to gross substitute valuations (GS). Second, we use techniques from learning theory to argue that the over-demand and welfare induced by a price vector converge to their expectations uniformly over the class of all price vectors, with respective sample complexity linear and quadratic in the number of goods in the market. These results make no assumption on the form of the valuation functions. These two results imply that under a mild genericity condition, the exact Walrasian equilibrium prices computed in a market are guaranteed to induce both low over-demand and high welfare when used in a new market where agents are sampled independently from the same distribution, whenever the number of agents is larger than the number of commodities in the market.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116614833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We analyse parallel algorithms in the context of exhaustive search over totally ordered sets. Imagine an infinite list of “boxes”, with a “treasure” hidden in one of them, where the boxes’ order reflects the importance of finding the treasure in a given box. At each time step, a search protocol executed by a searcher has the ability to peek into one box, and see whether the treasure is present or not. Clearly, the best strategy of a single searcher would be to open the boxes one by one, in increasing order. Moreover, by equally dividing the workload between them, k searchers can trivially find the treasure k times faster than one searcher. However, this straightforward strategy is very sensitive to failures (e.g., crashes of processors), and overcoming this issue seems to require a large amount of communication. We therefore address the question of designing parallel search algorithms maximizing their speed-up and maintaining high levels of robustness, while minimizing the amount of resources for coordination. Based on the observation that algorithms that avoid communication are inherently robust, we focus our attention on identifying the best running time performance of non-coordinating algorithms. Specifically, we devise non-coordinating algorithms that achieve a speed-up of 9/8 for two searchers, a speed-up of 4/3 for three searchers, and in general, a speed-up of k/4(1+1/k)2 for any k≥ 1 searchers. Thus, asymptotically, the speed-up is only four times worse compared to the case of full coordination. Moreover, these bounds are tight in a strong sense as no non-coordinating search algorithm can achieve better speed-ups. Our algorithms are surprisingly simple and hence applicable. However they are memory intensive and so we suggest a practical, memory efficient version, with a speed-up of (k2 − 1)/4k. That is, it is only a factor of (k+1)/(k−1) slower than the optimal algorithm. Overall, we highlight that, in faulty contexts in which coordination between the searchers is technically difficult to implement, intrusive with respect to privacy, and/or costly in term of resources, it might well be worth giving up on coordination, and simply run our non-coordinating exhaustive search algorithms.
{"title":"Parallel exhaustive search without coordination","authors":"P. Fraigniaud, Amos Korman, Yoav Rodeh","doi":"10.1145/2897518.2897541","DOIUrl":"https://doi.org/10.1145/2897518.2897541","url":null,"abstract":"We analyse parallel algorithms in the context of exhaustive search over totally ordered sets. Imagine an infinite list of “boxes”, with a “treasure” hidden in one of them, where the boxes’ order reflects the importance of finding the treasure in a given box. At each time step, a search protocol executed by a searcher has the ability to peek into one box, and see whether the treasure is present or not. Clearly, the best strategy of a single searcher would be to open the boxes one by one, in increasing order. Moreover, by equally dividing the workload between them, k searchers can trivially find the treasure k times faster than one searcher. However, this straightforward strategy is very sensitive to failures (e.g., crashes of processors), and overcoming this issue seems to require a large amount of communication. We therefore address the question of designing parallel search algorithms maximizing their speed-up and maintaining high levels of robustness, while minimizing the amount of resources for coordination. Based on the observation that algorithms that avoid communication are inherently robust, we focus our attention on identifying the best running time performance of non-coordinating algorithms. Specifically, we devise non-coordinating algorithms that achieve a speed-up of 9/8 for two searchers, a speed-up of 4/3 for three searchers, and in general, a speed-up of k/4(1+1/k)2 for any k≥ 1 searchers. Thus, asymptotically, the speed-up is only four times worse compared to the case of full coordination. Moreover, these bounds are tight in a strong sense as no non-coordinating search algorithm can achieve better speed-ups. Our algorithms are surprisingly simple and hence applicable. However they are memory intensive and so we suggest a practical, memory efficient version, with a speed-up of (k2 − 1)/4k. That is, it is only a factor of (k+1)/(k−1) slower than the optimal algorithm. Overall, we highlight that, in faulty contexts in which coordination between the searchers is technically difficult to implement, intrusive with respect to privacy, and/or costly in term of resources, it might well be worth giving up on coordination, and simply run our non-coordinating exhaustive search algorithms.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122725654","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
V. Braverman, Stephen R. Chestnut, Nikita Ivkin, David P. Woodruff
Given a stream p1, …, pm of items from a universe U, which, without loss of generality we identify with the set of integers {1, 2, …, n}, we consider the problem of returning all ℓ2-heavy hitters, i.e., those items j for which fj ≥ є √F2, where fj is the number of occurrences of item j in the stream, and F2 = ∑i ∈ [n] fi2. Such a guarantee is considerably stronger than the ℓ1-guarantee, which finds those j for which fj ≥ є m. In 2002, Charikar, Chen, and Farach-Colton suggested the CountSketch data structure, which finds all such j using Θ(log2 n) bits of space (for constant є > 0). The only known lower bound is Ω(logn) bits of space, which comes from the need to specify the identities of the items found. In this paper we show one can achieve O(logn loglogn) bits of space for this problem. Our techniques, based on Gaussian processes, lead to a number of other new results for data streams, including: (1) The first algorithm for estimating F2 simultaneously at all points in a stream using only O(lognloglogn) bits of space, improving a natural union bound. (2) A way to estimate the ℓ∞ norm of a stream up to additive error є √F2 with O(lognloglogn) bits of space, resolving Open Question 3 from the IITK 2006 list for insertion only streams.
{"title":"Beating CountSketch for heavy hitters in insertion streams","authors":"V. Braverman, Stephen R. Chestnut, Nikita Ivkin, David P. Woodruff","doi":"10.1145/2897518.2897558","DOIUrl":"https://doi.org/10.1145/2897518.2897558","url":null,"abstract":"Given a stream p1, …, pm of items from a universe U, which, without loss of generality we identify with the set of integers {1, 2, …, n}, we consider the problem of returning all ℓ2-heavy hitters, i.e., those items j for which fj ≥ є √F2, where fj is the number of occurrences of item j in the stream, and F2 = ∑i ∈ [n] fi2. Such a guarantee is considerably stronger than the ℓ1-guarantee, which finds those j for which fj ≥ є m. In 2002, Charikar, Chen, and Farach-Colton suggested the CountSketch data structure, which finds all such j using Θ(log2 n) bits of space (for constant є > 0). The only known lower bound is Ω(logn) bits of space, which comes from the need to specify the identities of the items found. In this paper we show one can achieve O(logn loglogn) bits of space for this problem. Our techniques, based on Gaussian processes, lead to a number of other new results for data streams, including: (1) The first algorithm for estimating F2 simultaneously at all points in a stream using only O(lognloglogn) bits of space, improving a natural union bound. (2) A way to estimate the ℓ∞ norm of a stream up to additive error є √F2 with O(lognloglogn) bits of space, resolving Open Question 3 from the IITK 2006 list for insertion only streams.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127945800","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The development of algorithms for hierarchical clustering has been hampered by a shortage of precise objective functions. To help address this situation, we introduce a simple cost function on hierarchies over a set of points, given pairwise similarities between those points. We show that this criterion behaves sensibly in canonical instances and that it admits a top-down construction procedure with a provably good approximation ratio.
{"title":"A cost function for similarity-based hierarchical clustering","authors":"S. Dasgupta","doi":"10.1145/2897518.2897527","DOIUrl":"https://doi.org/10.1145/2897518.2897527","url":null,"abstract":"The development of algorithms for hierarchical clustering has been hampered by a shortage of precise objective functions. To help address this situation, we introduce a simple cost function on hierarchies over a set of points, given pairwise similarities between those points. We show that this criterion behaves sensibly in canonical instances and that it admits a top-down construction procedure with a provably good approximation ratio.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130837162","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In recent years a high dimensional theory of expanders has emerged. The notion of combinatorial expansion of graphs (i.e. the Cheeger constant of a graph) has seen two generalizations to high dimensional simplicial complexes. One generalization, known as coboundary expansion, is due to Linial and Meshulem; the other, which we term here cosystolic expansion, is due to Gromov, who showed that cosystolic expanders have the topological overlapping property. No construction (either random or explicit) of bounded degree combinational expanders (according to either definition) were known until a recent work of Kaufman, Kazhdan and Lubotzky, which provided the first bounded degree cosystolic expanders of dimension two. No bounded degree combinatorial expanders are known in higher dimensions. In this work we present explicit bounded degree cosystolic expanders of every dimension. This solves affirmatively an open question raised by Gromov, who asked whether there exist bounded degree complexes with the topological overlapping property in every dimension. Moreover, we provide a local to global criterion on a complex that implies cosystolic expansion: Namely, for a d-dimensional complex, X, if its underlying graph is a good expander, and all its links are both coboundary expanders and good expander graphs, then the (d-1)-dimensional skeleton of the complex is a cosystolic expander.
{"title":"Bounded degree cosystolic expanders of every dimension","authors":"Shai Evra, T. Kaufman","doi":"10.1145/2897518.2897543","DOIUrl":"https://doi.org/10.1145/2897518.2897543","url":null,"abstract":"In recent years a high dimensional theory of expanders has emerged. The notion of combinatorial expansion of graphs (i.e. the Cheeger constant of a graph) has seen two generalizations to high dimensional simplicial complexes. One generalization, known as coboundary expansion, is due to Linial and Meshulem; the other, which we term here cosystolic expansion, is due to Gromov, who showed that cosystolic expanders have the topological overlapping property. No construction (either random or explicit) of bounded degree combinational expanders (according to either definition) were known until a recent work of Kaufman, Kazhdan and Lubotzky, which provided the first bounded degree cosystolic expanders of dimension two. No bounded degree combinatorial expanders are known in higher dimensions. In this work we present explicit bounded degree cosystolic expanders of every dimension. This solves affirmatively an open question raised by Gromov, who asked whether there exist bounded degree complexes with the topological overlapping property in every dimension. Moreover, we provide a local to global criterion on a complex that implies cosystolic expansion: Namely, for a d-dimensional complex, X, if its underlying graph is a good expander, and all its links are both coboundary expanders and good expander graphs, then the (d-1)-dimensional skeleton of the complex is a cosystolic expander.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126464372","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We show that any depth-d circuit for determining whether an n-node graph has an s-to-t path of length at most k must have size nΩ(k1/d/d) when k(n) ≤ n1/5, and nΩ(k1/5d/d) when k(n)≤ n. The previous best circuit size lower bounds were nkexp(−O(d)) (by Beame, Impagliazzo, and Pitassi (Computational Complexity 1998)) and nΩ((logk)/d) (following from a recent formula size lower bound of Rossman (STOC 2014)). Our lower bound is quite close to optimal, as a simple construction gives depth-d circuits of size nO(k2/d) for this problem (and strengthening our bound even to nkΩ(1/d) would require proving that undirected connectivity is not in NC1). Our proof is by reduction to a new lower bound on the size of small-depth circuits computing a skewed variant of the “Sipser functions” that have played an important role in classical circuit lower bounds. A key ingredient in our proof of the required lower bound for these Sipser-like functions is the use of random projections, an extension of random restrictions which were recently employed by Rossman, Servedio, and Tan (FOCS 2015). Random projections allow us to obtain sharper quantitative bounds while employing simpler arguments, both conceptually and technically, than in the previous works.
{"title":"Near-optimal small-depth lower bounds for small distance connectivity","authors":"Xi Chen, I. Oliveira, R. Servedio, Li-Yang Tan","doi":"10.1145/2897518.2897534","DOIUrl":"https://doi.org/10.1145/2897518.2897534","url":null,"abstract":"We show that any depth-d circuit for determining whether an n-node graph has an s-to-t path of length at most k must have size nΩ(k1/d/d) when k(n) ≤ n1/5, and nΩ(k1/5d/d) when k(n)≤ n. The previous best circuit size lower bounds were nkexp(−O(d)) (by Beame, Impagliazzo, and Pitassi (Computational Complexity 1998)) and nΩ((logk)/d) (following from a recent formula size lower bound of Rossman (STOC 2014)). Our lower bound is quite close to optimal, as a simple construction gives depth-d circuits of size nO(k2/d) for this problem (and strengthening our bound even to nkΩ(1/d) would require proving that undirected connectivity is not in NC1). Our proof is by reduction to a new lower bound on the size of small-depth circuits computing a skewed variant of the “Sipser functions” that have played an important role in classical circuit lower bounds. A key ingredient in our proof of the required lower bound for these Sipser-like functions is the use of random projections, an extension of random restrictions which were recently employed by Rossman, Servedio, and Tan (FOCS 2015). Random projections allow us to obtain sharper quantitative bounds while employing simpler arguments, both conceptually and technically, than in the previous works.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-09-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123395508","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We consider the well-studied cake cutting problem in which the goal is to identify an envy-free allocation based on a minimal number of queries from the agents. The problem has attracted considerable attention within various branches of computer science, mathematics, and economics. Although, the elegant Selfridge-Conway envy-free protocol for three agents has been known since 1960, it has been a major open problem to obtain a bounded envy-free protocol for more than three agents. The problem has been termed the central open problem in cake cutting. We solve this problem by proposing a discrete and bounded envy-free protocol for four agents.
{"title":"A discrete and bounded envy-free cake cutting protocol for four agents","authors":"H. Aziz, Simon Mackenzie","doi":"10.1145/2897518.2897522","DOIUrl":"https://doi.org/10.1145/2897518.2897522","url":null,"abstract":"We consider the well-studied cake cutting problem in which the goal is to identify an envy-free allocation based on a minimal number of queries from the agents. The problem has attracted considerable attention within various branches of computer science, mathematics, and economics. Although, the elegant Selfridge-Conway envy-free protocol for three agents has been known since 1960, it has been a major open problem to obtain a bounded envy-free protocol for more than three agents. The problem has been termed the central open problem in cake cutting. We solve this problem by proposing a discrete and bounded envy-free protocol for four agents.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122263570","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In the quantum state tomography problem, one wishes to estimate an unknown d-dimensional mixed quantum state ρ, given few copies. We show that O(d/ε) copies suffice to obtain an estimate ρ that satisfies ||ρ − ρ||F2 ≤ ε (with high probability). An immediate consequence is that O((ρ) · d/ε2) ≤ O(d2/ε2) copies suffice to obtain an ε-accurate estimate in the standard trace distance. This improves on the best known prior result of O(d3/ε2) copies for full tomography, and even on the best known prior result of O(d2log(d/ε)/ε2) copies for spectrum estimation. Our result is the first to show that nontrivial tomography can be obtained using a number of copies that is just linear in the dimension. Next, we generalize these results to show that one can perform efficient principal component analysis on ρ. Our main result is that O(k d/ε2) copies suffice to output a rank-k approximation ρ whose trace-distance error is at most ε more than that of the best rank-k approximator to ρ. This subsumes our above trace distance tomography result and generalizes it to the case when ρ is not guaranteed to be of low rank. A key part of the proof is the analogous generalization of our spectrum-learning results: we show that the largest k eigenvalues of ρ can be estimated to trace-distance error ε using O(k2/ε2) copies. In turn, this result relies on a new coupling theorem concerning the Robinson–Schensted–Knuth algorithm that should be of independent combinatorial interest.
{"title":"Efficient quantum tomography","authors":"R. O'Donnell, John Wright","doi":"10.1145/2897518.2897544","DOIUrl":"https://doi.org/10.1145/2897518.2897544","url":null,"abstract":"In the quantum state tomography problem, one wishes to estimate an unknown d-dimensional mixed quantum state ρ, given few copies. We show that O(d/ε) copies suffice to obtain an estimate ρ that satisfies ||ρ − ρ||F2 ≤ ε (with high probability). An immediate consequence is that O((ρ) · d/ε2) ≤ O(d2/ε2) copies suffice to obtain an ε-accurate estimate in the standard trace distance. This improves on the best known prior result of O(d3/ε2) copies for full tomography, and even on the best known prior result of O(d2log(d/ε)/ε2) copies for spectrum estimation. Our result is the first to show that nontrivial tomography can be obtained using a number of copies that is just linear in the dimension. Next, we generalize these results to show that one can perform efficient principal component analysis on ρ. Our main result is that O(k d/ε2) copies suffice to output a rank-k approximation ρ whose trace-distance error is at most ε more than that of the best rank-k approximator to ρ. This subsumes our above trace distance tomography result and generalizes it to the case when ρ is not guaranteed to be of low rank. A key part of the proof is the analogous generalization of our spectrum-learning results: we show that the largest k eigenvalues of ρ can be estimated to trace-distance error ε using O(k2/ε2) copies. In turn, this result relies on a new coupling theorem concerning the Robinson–Schensted–Knuth algorithm that should be of independent combinatorial interest.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125180257","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. Braverman, A. Garg, Tengyu Ma, Huy L. Nguyen, David P. Woodruff
We study the tradeoff between the statistical error and communication cost of distributed statistical estimation problems in high dimensions. In the distributed sparse Gaussian mean estimation problem, each of the m machines receives n data points from a d-dimensional Gaussian distribution with unknown mean θ which is promised to be k-sparse. The machines communicate by message passing and aim to estimate the mean θ. We provide a tight (up to logarithmic factors) tradeoff between the estimation error and the number of bits communicated between the machines. This directly leads to a lower bound for the distributed sparse linear regression problem: to achieve the statistical minimax error, the total communication is at least Ω(min{n,d}m), where n is the number of observations that each machine receives and d is the ambient dimension. These lower results improve upon Shamir (NIPS'14) and Steinhardt-Duchi (COLT'15) by allowing multi-round iterative communication model. We also give the first optimal simultaneous protocol in the dense case for mean estimation. As our main technique, we prove a distributed data processing inequality, as a generalization of usual data processing inequalities, which might be of independent interest and useful for other problems.
{"title":"Communication lower bounds for statistical estimation problems via a distributed data processing inequality","authors":"M. Braverman, A. Garg, Tengyu Ma, Huy L. Nguyen, David P. Woodruff","doi":"10.1145/2897518.2897582","DOIUrl":"https://doi.org/10.1145/2897518.2897582","url":null,"abstract":"We study the tradeoff between the statistical error and communication cost of distributed statistical estimation problems in high dimensions. In the distributed sparse Gaussian mean estimation problem, each of the m machines receives n data points from a d-dimensional Gaussian distribution with unknown mean θ which is promised to be k-sparse. The machines communicate by message passing and aim to estimate the mean θ. We provide a tight (up to logarithmic factors) tradeoff between the estimation error and the number of bits communicated between the machines. This directly leads to a lower bound for the distributed sparse linear regression problem: to achieve the statistical minimax error, the total communication is at least Ω(min{n,d}m), where n is the number of observations that each machine receives and d is the ambient dimension. These lower results improve upon Shamir (NIPS'14) and Steinhardt-Duchi (COLT'15) by allowing multi-round iterative communication model. We also give the first optimal simultaneous protocol in the dense case for mean estimation. As our main technique, we prove a distributed data processing inequality, as a generalization of usual data processing inequalities, which might be of independent interest and useful for other problems.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125187816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A. Ambainis, K. Balodis, Aleksandrs Belovs, Troy Lee, M. Santha, Juris Smotrovs
In 1986, Saks and Wigderson conjectured that the largest separation between deterministic and zero-error randomized query complexity for a total boolean function is given by the function f on n=2k bits defined by a complete binary tree of NAND gates of depth k, which achieves R0(f) = O(D(f)0.7537…). We show this is false by giving an example of a total boolean function f on n bits whose deterministic query complexity is Ω(n/log(n)) while its zero-error randomized query complexity is Õ(√n). We further show that the quantum query complexity of the same function is Õ(n1/4), giving the first example of a total function with a super-quadratic gap between its quantum and deterministic query complexities. We also construct a total boolean function g on n variables that has zero-error randomized query complexity Ω(n/log(n)) and bounded-error randomized query complexity R(g) = Õ(√n). This is the first super-linear separation between these two complexity measures. The exact quantum query complexity of the same function is QE(g) = Õ(√n). These functions show that the relations D(f) = O(R1(f)2) and R0(f) = Õ(R(f)2) are optimal, up to poly-logarithmic factors. Further variations of these functions give additional separations between other query complexity measures: a cubic separation between Q and R0, a 3/2-power separation between QE and R, and a 4th power separation between approximate degree and bounded-error randomized query complexity. All of these examples are variants of a function recently introduced by Goos, Pitassi, and Watson which they used to separate the unambiguous 1-certificate complexity from deterministic query complexity and to resolve the famous Clique versus Independent Set problem in communication complexity.
{"title":"Separations in query complexity based on pointer functions","authors":"A. Ambainis, K. Balodis, Aleksandrs Belovs, Troy Lee, M. Santha, Juris Smotrovs","doi":"10.1145/2897518.2897524","DOIUrl":"https://doi.org/10.1145/2897518.2897524","url":null,"abstract":"In 1986, Saks and Wigderson conjectured that the largest separation between deterministic and zero-error randomized query complexity for a total boolean function is given by the function f on n=2k bits defined by a complete binary tree of NAND gates of depth k, which achieves R0(f) = O(D(f)0.7537…). We show this is false by giving an example of a total boolean function f on n bits whose deterministic query complexity is Ω(n/log(n)) while its zero-error randomized query complexity is Õ(√n). We further show that the quantum query complexity of the same function is Õ(n1/4), giving the first example of a total function with a super-quadratic gap between its quantum and deterministic query complexities. We also construct a total boolean function g on n variables that has zero-error randomized query complexity Ω(n/log(n)) and bounded-error randomized query complexity R(g) = Õ(√n). This is the first super-linear separation between these two complexity measures. The exact quantum query complexity of the same function is QE(g) = Õ(√n). These functions show that the relations D(f) = O(R1(f)2) and R0(f) = Õ(R(f)2) are optimal, up to poly-logarithmic factors. Further variations of these functions give additional separations between other query complexity measures: a cubic separation between Q and R0, a 3/2-power separation between QE and R, and a 4th power separation between approximate degree and bounded-error randomized query complexity. All of these examples are variants of a function recently introduced by Goos, Pitassi, and Watson which they used to separate the unambiguous 1-certificate complexity from deterministic query complexity and to resolve the famous Clique versus Independent Set problem in communication complexity.","PeriodicalId":442965,"journal":{"name":"Proceedings of the forty-eighth annual ACM symposium on Theory of Computing","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116884504","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：