1/17 Deterministic Discrepancy Minimization Nikhil Bansal (TU Eindhoven) Joel Spencer (NYU)

Slides:



Advertisements
Similar presentations
Parikshit Gopalan Georgia Institute of Technology Atlanta, Georgia, USA.
Advertisements

Subspace Embeddings for the L1 norm with Applications Christian Sohler David Woodruff TU Dortmund IBM Almaden.
Fast Johnson-Lindenstrauss Transform(s) Nir Ailon Edo Liberty, Bernard Chazelle Bertinoro Workshop on Sublinear Algorithms May 2011.
The Contest between Simplicity and Efficiency in Asynchronous Byzantine Agreement Allison Lewko The University of Texas at Austin TexPoint fonts used in.
Submodular Set Function Maximization via the Multilinear Relaxation & Dependent Rounding Chandra Chekuri Univ. of Illinois, Urbana-Champaign.
Time-Space Tradeoffs in Resolution: Superpolynomial Lower Bounds for Superlinear Space Chris Beck Princeton University Joint work with Paul Beame & Russell.
Satyen Kale (Yahoo! Research) Joint work with Sanjeev Arora (Princeton)
Discrepancy and SDPs Nikhil Bansal (TU Eindhoven, Netherlands ) August 24, ISMP 2012, Berlin.
An Approximate Truthful Mechanism for Combinatorial Auctions An Internet Mathematics paper by Aaron Archer, Christos Papadimitriou, Kunal Talwar and Éva.
List decoding Reed-Muller codes up to minimal distance: Structure and pseudo- randomness in coding theory Abhishek Bhowmick (UT Austin) Shachar Lovett.
1/26 Constructive Algorithms for Discrepancy Minimization Nikhil Bansal (IBM)
Discrepancy Minimization by Walking on the Edges Raghu Meka (IAS/DIMACS) Shachar Lovett (IAS)
The Randomization Repertoire Rajmohan Rajaraman Northeastern University, Boston May 2012 Chennai Network Optimization WorkshopThe Randomization Repertoire1.
Heuristics for the Hidden Clique Problem Robert Krauthgamer (IBM Almaden) Joint work with Uri Feige (Weizmann)
Approximation Algorithms for Capacitated Set Cover Ravishankar Krishnaswamy (joint work with Nikhil Bansal and Barna Saha)
1 Learning with continuous experts using Drifting Games work with Robert E. Schapire Princeton University work with Robert E. Schapire Princeton University.
C&O 355 Mathematical Programming Fall 2010 Lecture 20 N. Harvey TexPoint fonts used in EMF. Read the TexPoint manual before you delete this box.: AA A.
Dependent Randomized Rounding in Matroid Polytopes (& Related Results) Chandra Chekuri Jan VondrakRico Zenklusen Univ. of Illinois IBM ResearchMIT.
How should we define corner points? Under any reasonable definition, point x should be considered a corner point x What is a corner point?
Noga Alon Institute for Advanced Study and Tel Aviv University
1/17 Optimal Long Test with One Free Bit Nikhil Bansal (IBM) Subhash Khot (NYU)
A Randomized Polynomial-Time Simplex Algorithm for Linear Programming Daniel A. Spielman, Yale Joint work with Jonathan Kelner, M.I.T.
1 List Coloring and Euclidean Ramsey Theory TexPoint fonts used in EMF. Read the TexPoint manual before you delete this box.: A A A A Noga Alon, Tel Aviv.
Approximation Algoirthms: Semidefinite Programming Lecture 19: Mar 22.
Semidefinite Programming
1 Introduction to Linear and Integer Programming Lecture 9: Feb 14.
1 Optimization problems such as MAXSAT, MIN NODE COVER, MAX INDEPENDENT SET, MAX CLIQUE, MIN SET COVER, TSP, KNAPSACK, BINPACKING do not have a polynomial.
Rules for means Rule 1: If X is a random variable and a and b are fixed numbers, then Rule 2: If X and Y are random variables, then.
Approximate Nearest Neighbors and the Fast Johnson-Lindenstrauss Transform Nir Ailon, Bernard Chazelle (Princeton University)
Michael Bender - SUNY Stony Brook Dana Ron - Tel Aviv University Testing Acyclicity of Directed Graphs in Sublinear Time.
Theta Function Lecture 24: Apr 18. Error Detection Code Given a noisy channel, and a finite alphabet V, and certain pairs that can be confounded, the.
Prof. Bart Selman Module Probability --- Part e)
Scalable Information-Driven Sensor Querying and Routing for ad hoc Heterogeneous Sensor Networks Maurice Chu, Horst Haussecker and Feng Zhao Xerox Palo.
Daniel Kroening and Ofer Strichman Decision Procedures An Algorithmic Point of View Deciding ILPs with Branch & Bound ILP References: ‘Integer Programming’
Discrepancy Minimization by Walking on the Edges Raghu Meka (IAS & DIMACS) Shachar Lovett (IAS)
1/24 Algorithms for Generalized Caching Nikhil Bansal IBM Research Niv Buchbinder Open Univ. Israel Seffi Naor Technion.
1/30 Discrepancy and SDPs Nikhil Bansal (TU Eindhoven)
Beating the Union Bound by Geometric Techniques Raghu Meka (IAS & DIMACS)
Decision Procedures An Algorithmic Point of View
Complexity Theory Lecture 2 Lecturer: Moni Naor. Recap of last week Computational Complexity Theory: What, Why and How Overview: Turing Machines, Church-Turing.
Correlation testing for affine invariant properties on Shachar Lovett Institute for Advanced Study Joint with Hamed Hatami (McGill)
CS774. Markov Random Field : Theory and Application Lecture 08 Kyomin Jung KAIST Sep
Weighted Geometric Set Multicover via Quasi-uniform Sampling (ESA 2012) Kirk Pruhs (U. Pittsburgh) Coauthor: Nikhil Bansal (TU Eindhoven)
© 2009 IBM Corporation 1 Improving Consolidation of Virtual Machines with Risk-aware Bandwidth Oversubscription in Compute Clouds Amir Epstein Joint work.
Edge-disjoint induced subgraphs with given minimum degree Raphael Yuster 2012.
PODC Distributed Computation of the Mode Fabian Kuhn Thomas Locher ETH Zurich, Switzerland Stefan Schmid TU Munich, Germany TexPoint fonts used in.
Testing the independence number of hypergraphs
ME451 Kinematics and Dynamics of Machine Systems Review of Differential Calculus 2.5, 2.6 September 11, 2013 Radu Serban University of Wisconsin-Madison.
1/19 Minimizing weighted completion time with precedence constraints Nikhil Bansal (IBM) Subhash Khot (NYU)
1/30 Constructive Algorithms for Discrepancy Minimization Nikhil Bansal (IBM)
Dimension Reduction using Rademacher Series on Dual BCH Codes Nir Ailon Edo Liberty.
Unique Games Approximation Amit Weinstein Complexity Seminar, Fall 2006 Based on: “Near Optimal Algorithms for Unique Games" by M. Charikar, K. Makarychev,
Random Geometric Graph Model Model for ad hoc/sensor networks n nodes placed in d-dim space Connectivity threshold r Two nodes u,v connected iff ||u-v||
Pseudo-random generators Talk for Amnon ’ s seminar.
Technion Haifa Research Labs Israel Institute of Technology Underapproximation for Model-Checking Based on Random Cryptographic Constructions Arie Matsliah.
The NP class. NP-completeness
Umans Complexity Theory Lectures
Lap Chi Lau we will only use slides 4 to 19
Topics in Algorithms Lap Chi Lau.
Fast Dimension Reduction MMDS 2008
CS154, Lecture 18:.
Dana Moshkovitz The Institute For Advanced Study
Chapter 5. Optimal Matchings
Structural Properties of Low Threshold Rank Graphs
Faster Space-Efficient Algorithms for Subset Sum
Discrepancy and Optimization
On Approximating Covering Integer Programs
Online Algorithms via Projections set cover, paging, k-server
Clustering.
Sublinear Algorihms for Big Data
Presentation transcript:

1/17 Deterministic Discrepancy Minimization Nikhil Bansal (TU Eindhoven) Joel Spencer (NYU)

2/17 Combinatorial Discrepancy Universe: U= [1,…,n] Subsets: S 1,S 2,…,S m Problem: Color elements red/blue so each subset is colored as evenly as possible. CS: Computational Geometry, Comb. Optimization, Monte-Carlo simulation, Machine learning, Complexity, Pseudo-Randomness, … Math: Dynamical Systems, Combinatorics, Mathematical Finance, Number Theory, Ramsey Theory, Algebra, Measure Theory, … S1S1 S2S2 S3S3 S4S4

3/17 General Set System Universe: U= [1,…,n] Subsets: S 1,S 2,…,S m Find  : [n] ! {-1,+1} to Minimize |  (S)| 1 = max S |  i 2 S  (i) | For simplicity consider m=n henceforth.

4/17 Simple Algorithm Random: Color each element i independently as x(i) = +1 or -1 with probability ½ each. Thm: Discrepancy = O (n log n) 1/2 Pf: For each set, expect O(n 1/2 ) discrepancy Standard tail bounds: Pr[ |  i 2 S x(i) | ¸ c n 1/2 ] ¼ e -c 2 Union bound + Choose c ¼ (log n) 1/2 Analysis tight: Random actually incurs  (n log n) 1/2 ).

5/17 Better Colorings Exist! [Spencer 85]: (Six standard deviations suffice) Always exists coloring with discrepancy · 6n 1/2 Tight: Cannot beat 0.5 n 1/2 (Hadamard Matrix, “orthogonal” sets) Inherently non-constructive proof (pigeonhole principle on exponentially large universe) Challenge: Can we find it algorithmically ? (Certain algorithms do not work) Conjecture [Alon-Spencer]: May not be possible.

6/17 Algorithmic Results [Bansal 10]: Efficient (randomized) algorithm for Spencer’s result. Technique: SDPs (new rounding idea) Use several SDPs over time (guided by the non-constructive method). General: Geometric problems, Beck Fiala setting, k-permutation problem, pseudo-approximation of discrepancy, … Thm: Deterministic Algorithm for Spencer’s (and other) results.

This Talk A 7/17

Derandomizing Chernoff (Pessimistic estimators, exp. moments, hyp. cosine rule, …) 8/17

The Problem 9/17

10/17 Algorithm (at high level) Cube: {-1,1} n start finish Each dimension: A variable Each vertex: A rounding

11/17 SDP relaxations

12/17 Analysis (at high level) Cube: {-1,1} n Analysis: Progress: Few steps to reach a vertex (walk has high variance) Low Discrepancy: For each equation, discrepancy random walk has low variance start finish Each dimension: An Element Each vertex: A Coloring

Making it Deterministic Need to find an update that i)Makes Progress ii)Adds low discrepancy to equations. Recall, for Chernoff: Round one variable at a time (Progress) Whether -1 or +1, guided by the potential. (Low Discrepancy) 13/17

Tracking the properties 14/17

Our fix 15/17 origin x(t-1) x(t): New position

Trouble 16/17

17/17 Concluding Remarks Idea: Add new constraints to force a deterministic choice to exist. Works more generally for other discrepancy problems. Can potentially have other applications. Thank You!

Techniques 18/17 Entropy Method Spencer’s Result Bansal’s Result SDPs New “orthogonality” idea (based on entropy) + K-wise independence, pessimistic estimators, … This Result