pairwise comparisons with flexible time dynamics
play

Pairwise Comparisons with Flexible Time-Dynamics Lucas Maystre , - PowerPoint PPT Presentation

May 30, 2023 •366 likes •554 views

Pairwise Comparisons with Flexible Time-Dynamics Lucas Maystre , Victor Kristof, Matthias Grossglauser KDD Research Track 2 August 6 th , 2019 Pairwise comparison data Association football example: Questions we might want to ask: Team 1

  1. Pairwise Comparisons with Flexible Time-Dynamics Lucas Maystre , Victor Kristof, Matthias Grossglauser KDD Research Track 2 — August 6 th , 2019

  2. Pairwise comparison data Association football example: Questions we might want to ask: Team 1 Team 2 Score “How can we quantify the skill of France?” France Portugal 2-5 Luxembourg Greece 0-3 “How likely is South Korea ... ... ... to beat Germany?” Turkey Slovakia 1-1 Bulgaria Kosovo 0-1 → we need pairwise comparison models. 2

  3. <latexit sha1_base64="UP3JbDQ68thUsUruLlvz9QvaE80=">ACAXicbVDLSgMxFL1TX7W+qi7dBIvgqsxUQZcFRVxWsA+YDiWTpm1oMhmSjFCGrvwGt7p2J279Epf+iZl2Frb1QOBwzr3ckxPGnGnjut9OYW19Y3OruF3a2d3bPygfHrW0TBShTSK5VJ0Qa8pZRJuGU47saJYhJy2w/FN5refqNJMRo9mEtNA4GHEBoxgYyW/K7AZEczT2mvXHGr7gxolXg5qUCORq/80+1LkgaGcKx1r7nxiZIsTKMcDotdRNY0zGeEh9SyMsqA7SWeQpOrNKHw2ksi8yaKb+3Uix0HoiQjuZRdTLXib+5/mJGVwHKYvixNCIzA8NEo6MRNn/UZ8pSgyfWIKJYjYrIiOsMDG2pYUrocg68ZYbWCWtWtW7qNYeLiv1u7ydIpzAKZyDB1dQh3toQBMISHiBV3hznp1358P5nI8WnHznGBbgfP0CWKuX+A=</latexit> <latexit sha1_base64="HG+BMlUV/2K3oGVTANntbUCUhw=">ACKnicbVDLSsNAFJ34rPUVdelmsAgVsSRV0I1QcONKtgHJCFMpN27OTBzEQsIf/hR/gNbnXtrgS/BAnbRa29cAwh3PO5c4cL2ZUSMYa0vLK6tr6WN8ubW9s6uvrfFlHCMWnhiEW86yFBGA1JS1LJSDfmBAUeIx1veJP7nSfCBY3CBzmKiROgfkh9ipFUkqvX4yqFtkgwho8n8BraPkc4NbPUhKfQJs+xBc9gVbhUXcJVESdz9YpRMyaAi8QsSAUaLr6t92LcBKQUGKGhLBMI5ZOirikmJGsbCeCxAgPUZ9YioYoIMJ3/L4LFSetCPuDqhBP170SKAiFGgaeSAZIDMe/l4n+elUj/yklpGCeShHi6yE8YlBHMi4I9ygmWbKQIwpyqt0I8QKodqeqc2eIFeSfmfAOLpF2vme1+v1FpXFXtFMCh+AIVIEJLkED3ImaAEMXsAbeAcf2qv2qY21r2l0StmDsAMtJ9f16ykeA=</latexit> <latexit sha1_base64="iBqJEKBf+8hcXui8sJPkCiQKLc4=">ACG3icbVDLSgMxFM34rPVdekmWIQKUmaqoMuCG1elgn1AZyZNOGJpkhyQhlmA/wI/wGt7p2J25duPRPzLSzsK0HAodzuXeHD9iVGnb/rZWVtfWNzYLW8Xtnd29/dLBYVuFscSkhUMWyq6PFGFUkJampFuJAniPiMdf3yT+Z1HIhUNxb2eRMTjaChoQDHSRuqXyqpPoasohxC6HOkRixpBX7PFOHD3UzkzKrtpTwGXi5KQMcjT7pR93EOKYE6ExQ0r1HDvSXoKkpiRtOjGikQIj9GQ9AwViBPlJdPpPDUKAMYhNI8oeFU/TuRIK7UhPsmd2rFr1M/M/rxTq49hIqolgTgWeLgphBHcKsGTigkmDNJoYgLKm5FeIRkghr09/cFp+nphNnsYFl0q5VnYtq7e6yXG/k7RTAMTgBFeCAK1AHt6AJWgCDJ/ACXsGb9Wy9Wx/W5y6YuUzR2AO1tcvXbGgVQ=</latexit> <latexit sha1_base64="Ix18cYdMx7NrDq0ECn91v7Keh+Q=">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</latexit> [Zermelo, 1928] 
 Latent skill model [Thurstone, 1927] s i ∼ N (0 , σ 2 ) Germany 1.18 latent skill of i Brazil 1.09 France 0.68 1 p ( i � j ) = 1 + exp[ � ( s i � s j )] 0 “ i wins over j ” Switzerland -0.74 Given data , the posterior distribution is D Iceland -1.22 Y p ( s | D ) / p ( s ) p ( i � j ) ( i,j ) ∈ D 3

  4. Actual setting Data come with a timestamp . Questions we might want to ask: Date Team 1 Team 2 Score “How strong was France in 1972 ? In 2018 ?” 1923-09-15 France Portugal 2-5 1923-09-16 Luxembourg Greece 0-3 “How likely is South Korea ... ... ... ... to beat Germany today ?” 2018-06-21 Turkey Slovakia 1-1 2018-06-21 Bulgaria Kosovo 0-1 → we need dynamic models. 4

  5. <latexit sha1_base64="LuQOmJp/62XYXmAp4JLOqHRlnj8=">ACHicbVDLSgNBEJyNrxhfUY9eBqOYQAi7UdBjwIOeJIJ5QLKE2clsMmRmd5npFcKSH/Aj/AavevYmXgWP/omTx8EkFjQUVd10d3mR4Bps+9tKrayurW+kNzNb2zu7e9n9g7oOY0VZjYiVE2PaCZ4wGrAQbBmpBiRnmANb3A9huPTGkeBg8wjJgrS/gPqcEjNTJnugOz0MBtzWXuC0J9JVMbqjl3EgzwUMZwV3E42Z5fsCfAycWYkh2aodrI/7W5IY8kCoIJo3XLsCNyEKOBUsFGmHWsWETogPdYyNCSaTeZfDPCp0bpYj9UpgLAE/XvREKk1kPpmc7xvXrRG4v/ea0Y/Cs34UEUAwvodJEfCwhHkeDu1wxCmJoCKGKm1sx7RNFKJgA57Z4cmQycRYTWCb1csk5L5XvL3KVu1k6aXSEjlEeOegSVdAtqIaougJvaBX9GY9W+/Wh/U5bU1Zs5lDNAfr6xeNXqBd</latexit> <latexit sha1_base64="WbOUPImugAWxZeTwfab5BbunGSI=">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</latexit> This work: kickscore Skill becomes a (latent) stochastic process s i ( t ) ∼ GP[0 , k ( t, t 0 )] covariance function, defines time dynamics Obs. model is conditionally parametric : p ( i � j | t ) = 1 1 + exp { � [ s i ( t ) � s j ( t )] } 5

  6. Covariance functions Brownian motion Mean-reverting, stationary dynamics Smooth dynamics Discontinuities 6

  7. Outline 1 2 Inference Experimental algorithm evaluation 7

Recommend

Graph Resistance and Learning from Pairwise Comparisons pairwise comparisons of items. In

Graph Resistance and Learning from Pairwise Comparisons pairwise comparisons of items. In

Alex Olshevsky Department of ECE, Boston University Joint work with Julien Hendrickx (UC Louvain) and Venkatesh Saligrama (BU) Graph Resistance and Learning from Pairwise Comparisons pairwise comparisons of items. In many contexts,

759 views • 50 slides
Rank Aggregation from Pairwise Comparisons in the Presence of Adversarial Corruptions Arpit

Rank Aggregation from Pairwise Comparisons in the Presence of Adversarial Corruptions Arpit

Rank Aggregation from Pairwise Comparisons in the Presence of Adversarial Corruptions Arpit Agarwal, Shivani Agarwal, Sanjeev Khanna, Prathamesh Patil ICML 2020 Rank Aggregation from Pairwise Comparisons In many practical applications, the

844 views • 37 slides
Matrix Profile XIV: Scaling Time Series Motif Discovery with GPUs to Break a Quintillion Pairwise

Matrix Profile XIV: Scaling Time Series Motif Discovery with GPUs to Break a Quintillion Pairwise

Matrix Profile XIV: Scaling Time Series Motif Discovery with GPUs to Break a Quintillion Pairwise Comparisons a Day and Beyond Zachary Zimmerman, Kaveh Kamgar, Nader Shakibay Senobari, Yan Zhu, Brian Crites, Gareth Funning, Philip Brisk, Eamonn

917 views • 25 slides
Using Flexible Time Scale to Explore the Validity of Agent-based Models of Ecosystem Dynamics

Using Flexible Time Scale to Explore the Validity of Agent-based Models of Ecosystem Dynamics

Using Flexible Time Scale to Explore the Validity of Agent-based Models of Ecosystem Dynamics Application to Simulation of a Wild Rodent Population in a Changing Agricultural Landscape Jean Le Fur and Moussa Sall Simultech - Porto /

710 views • 45 slides
http://www.sztaki.mta.hu/%7Ebozoki/slides p. 1/65 Multi Criteria Decision Making Pairwise

http://www.sztaki.mta.hu/%7Ebozoki/slides p. 1/65 Multi Criteria Decision Making Pairwise

Weighting and ranking based on pairwise comparisons Pros sszehasonlts alap slyozs s rangsorols Sndor Bozki Institute for Computer Science and Control Hungarian Academy of Sciences (MTA SZTAKI) Corvinus University of

1.09k views • 65 slides
Small-scale galaxy dynamics: the pairwise velocity dispersion Jon Loveday University of Sussex

Small-scale galaxy dynamics: the pairwise velocity dispersion Jon Loveday University of Sussex

Small-scale galaxy dynamics: the pairwise velocity dispersion Jon Loveday University of Sussex Outline RSD overview Galaxy pairwise velocity dispersion (PVD) - why measure it? GAMA data and mocks Ways of measuring PVD Results:

494 views • 20 slides
BradleyTerry2: Flexible Models for Paired Comparisons Heather Turner and David Firth Department

BradleyTerry2: Flexible Models for Paired Comparisons Heather Turner and David Firth Department

BradleyTerry2: Flexible Models for Paired Comparisons Heather Turner and David Firth Department of Statistics University of Warwick, UK 21 July 2010 H. Turner and D. Firth (Warwick, UK) BradleyTerry2 useR! 2010 1 / 15 Pair Comparisons In

341 views • 15 slides
Possibilities for Possibilities for flexible time arrangements flexible time arrangements

Possibilities for Possibilities for flexible time arrangements flexible time arrangements

MODULE 2. MODULE 2. Possibilities for Possibilities for flexible time arrangements flexible time arrangements FACILITATING FAMILY LEARNING ON WORK &amp; LIFE BALANCE Project No: 502889-LLP-1-2009-1-LT-GRUNDTVIG-GMP MODULE 2. MODULE 2. AIM:

408 views • 15 slides
Multiple Comparisons Occasionally, e.g., at the start of a research project, we do not have a

Multiple Comparisons Occasionally, e.g., at the start of a research project, we do not have a

Lecture 7.1: Multiple Comparisons (A non-quiz topic) Examples of the need for multiple comparisons The problem with multiple comparisons post hoc; an outline of a solution Specific solutions: Fisher LSD, Tukey HSD, Holm (same as

840 views • 69 slides
Flexible multibody dynamics: From FE formulations to control and optimization Olivier Brls

Flexible multibody dynamics: From FE formulations to control and optimization Olivier Brls

Flexible multibody dynamics: From FE formulations to control and optimization Olivier Brls Multibody &amp; Mechatronic Systems Laboratory Department of Aerospace and Mechanical Engineering University of Lige, Belgium Acknowledgement to

1.22k views • 68 slides
Molecular dynamics flexible fitting (MDFF) James C. (JC) Gumbart Georgia Institute of

Molecular dynamics flexible fitting (MDFF) James C. (JC) Gumbart Georgia Institute of

Molecular dynamics flexible fitting (MDFF) James C. (JC) Gumbart Georgia Institute of Technology, Atlanta Computational Biophysics Workshop | DICP | July 12 2018 Structural biology continuum 1 cm 1 m 1 mm 1 nm 1 Light microscopy

502 views • 34 slides
Web Dynamics Part 1 - Introduction 1.1 Dimensions of dynamics in the Web 1.2 Application

Web Dynamics Part 1 - Introduction 1.1 Dimensions of dynamics in the Web 1.2 Application

Web Dynamics Part 1 - Introduction 1.1 Dimensions of dynamics in the Web 1.2 Application examples Summer Term 2010 Web Dynamics 1-1 Why Web Dynamics? From Wikipedia: In physics the term dynamics customarily refers to the time evolution of

440 views • 42 slides
Molecular Dynamics Flexible Fitting Ryan McGreevy Research Programmer University of Illinois at

Molecular Dynamics Flexible Fitting Ryan McGreevy Research Programmer University of Illinois at

Molecular Dynamics Flexible Fitting Ryan McGreevy Research Programmer University of Illinois at Urbana-Champaign NIH Resource for Macromolecular Modeling and Bioinformatics M olecular D ynamics F lexible F itting (Ribosome-bound YidC)

439 views • 14 slides
Web Dynamics Part 1 Introduction 1.1 Dimensions of dynamics in the Web 1.2 Application examples

Web Dynamics Part 1 Introduction 1.1 Dimensions of dynamics in the Web 1.2 Application examples

Web Dynamics Part 1 Introduction 1.1 Dimensions of dynamics in the Web 1.2 Application examples Summer Term 2009 Web Dynamics 1 1 Why Web Dynamics? From Wikipedia: In physics the term dynamics customarily refers to the time evolution of

608 views • 35 slides
Changes in what employees value Employers offer more frequently Flexible working hours 1

Changes in what employees value Employers offer more frequently Flexible working hours 1

Changes in what employees value Employers offer more frequently Flexible working hours 1 day of home office every week Daily commute time counted as work time Flexible working location Timelines promotion and international moves

419 views • 8 slides
Online Learning with Pairwise Loss Functions Online Learning with Pairwise Loss Functions MLSIG

Online Learning with Pairwise Loss Functions Online Learning with Pairwise Loss Functions MLSIG

Online Learning with Pairwise Loss Functions Online Learning with Pairwise Loss Functions MLSIG Seminar Series, Dept. of CSA, IISc Joint work with B. Sriperumbudur, P. Jain, H. Karnick Purushottam Kar MLO Group, Microsoft Research India Outline

464 views • 28 slides
Web Dynamics Part 5 Searching the Past 5.1 Time-travel problems 5.2 Efficient Time-Travel

Web Dynamics Part 5 Searching the Past 5.1 Time-travel problems 5.2 Efficient Time-Travel

Web Dynamics Part 5 Searching the Past 5.1 Time-travel problems 5.2 Efficient Time-Travel Search 5.3 Temporal measures of page importance Summer Term 2010 Web Dynamics 5-1 Time Travel Problems on the Web Search engines index only the

1.55k views • 47 slides
STRATEGIC UPDATE &amp; FLEXIBLE PATH Board Public Input Session J u n e 1 3 , 2 0 1 8 I n f o

STRATEGIC UPDATE &amp; FLEXIBLE PATH Board Public Input Session J u n e 1 3 , 2 0 1 8 I n f o

STRATEGIC UPDATE &amp; FLEXIBLE PATH Board Public Input Session J u n e 1 3 , 2 0 1 8 I n f o r m a t i o n a l U p d a t e 1 AGENDA OVERVIEW OPERATIONAL DYNAMICS SAVE FOR TOMORROW ENERGY PLAN (STEP) OUR FLEXIBLE PATH

582 views • 44 slides
Pairwise Sequence Alignment: Dynamic Programming Algorithms COMP 571 Luay Nakhleh, Rice

Pairwise Sequence Alignment: Dynamic Programming Algorithms COMP 571 Luay Nakhleh, Rice

1 Pairwise Sequence Alignment: Dynamic Programming Algorithms COMP 571 Luay Nakhleh, Rice University 2 DP Algorithms for Pairwise Alignment The number of all possible pairwise alignments (if gaps are allowed) is exponential in the length

386 views • 9 slides
Pairwise comparison, and other methods MATH 105: Contemporary Mathematics University of

Pairwise comparison, and other methods MATH 105: Contemporary Mathematics University of

Pairwise comparison, and other methods MATH 105: Contemporary Mathematics University of Louisville October 26, 2017 Pairwise Comparison 2 / 12 How to Simplify Voting Every method weve seen so far seems flawed somehow. And yet choosing

389 views • 6 slides
Pairwise Alignment Mark Voorhies 3/27/2012 Mark Voorhies Pairwise Alignment Review: Tips and

Pairwise Alignment Mark Voorhies 3/27/2012 Mark Voorhies Pairwise Alignment Review: Tips and

Pairwise Alignment Mark Voorhies 3/27/2012 Mark Voorhies Pairwise Alignment Review: Tips and tricks Making a file executable: chmod a+x pydotter . py Handling file/directory names with spaces: cd My \ D i r e c t o r y \ with \ Spaces

955 views • 57 slides
ENCOPLOT: Pairwise Sequence Matching in Linear Time Applied to Plagiarism Detection Cristian

ENCOPLOT: Pairwise Sequence Matching in Linear Time Applied to Plagiarism Detection Cristian

Network Security Plagiarism Encoplot ENCOPLOT: Pairwise Sequence Matching in Linear Time Applied to Plagiarism Detection Cristian Grozea, Christian Gehl , Marius N. Popescu* christian.gehl@first.fraunhofer.de Fraunhofer Institute FIRST (IDA)

337 views • 32 slides
Correction for multiple comparisons in FreeSurfer 1 Problem of Multiple Comparisons p &lt; 10 -7

Correction for multiple comparisons in FreeSurfer 1 Problem of Multiple Comparisons p &lt; 10 -7

Correction for multiple comparisons in FreeSurfer 1 Problem of Multiple Comparisons p &lt; 10 -7 p &lt; 0.01 p &lt; 0.10 p value is probability that a voxel is falsely activated Threshold too liberal: many false positives Threshold too

230 views • 21 slides
WP8 Molecular Dynamics of Time-Dependent Phenomena WP8 Molecular Dynamics of

WP8 Molecular Dynamics of Time-Dependent Phenomena WP8 Molecular Dynamics of

WP8 Molecular Dynamics of Time-Dependent Phenomena WP8 Molecular Dynamics of Time-Dependent Phenomena (Kirchner, Vhringer) (Kirchner, Vhringer) Kinetics of chemical reactions during Bachelor- and Master curricula (Kinetics and

428 views • 5 slides

More recommend