SlideShare a Scribd company logo
1 of 18
Download to read offline
[Summery of] Robust and Effective Metric Learning Using
Capped Trace Norm [1] (KDD 2016)
shiba44
Outline
Introduction
Background
Proposed Method
Experiment
Conclusion
0 / 14
Introduction
Metric Learning aims automatically learning from train data.
similar
dissimilar
Example: Learning distance of image
They introduce low-rank regularization for Mahalanobis distance metric learning.
dM (xi, xj)
√
(xi − xj)⊤M(xi − xj)
1 / 14
Background:Weak Constraints
Pairwise
similar
dissimilar
Tripletwise
similar
dissimilar
Quadrupletwise
similar
dissimilar
Pairwise: dM (xi, xj) is smaller, dM (xi′, xj′) is bigger.
Tripletwise: dM (xi, xj) is smaller than dM (xj, xk).
Quadrupletwise: dM (xi, xj) is smaller than dM (xk, xl).
2 / 14
Background:low-rank and existing method
M is low-rank → Mahalanobis distance is defined in low dim space.
Rank minimization is NP Hard [2] ⇒ Use approximation of that.
Trace Norm regularization
▶ Minimizes sum of all singular values σ(M).
Reg(M)
∑
s
σs(M).
Changing of one large singular value affects overall.
Fantope regularization
▶ Minimizes sum of k smallest singular values.
Reg(M)
k∑
s
σs(M).
Be sensitive for hyper-parameter k. 3 / 14
Proposed Method
In proposed method, use Capped Trace Norm regularization.
Capped Trace Norm regularization
Only minimizes the singular values that are smaller than ϵ.
Reg(M)
∑
s
min(σs(M),ϵ).
Reduce the effect of changing large singular value.
Stable for hyper-parameter than Fantope.
4 / 14
Proposed Method
Optimization Problem:
min
M ∈Sd
+
∑
(i,j,k,l)∈A
[
δijkl +
⟨
M, xij x⊤
ij − xkl x⊤
kl
⟩]
+
Degree of violation for quadrupletwise constraints
+
γ
2
∑
s
min(σs(M),ϵ)
Regularization term
,where A =
{
(i, j, k,l) : dM (xk, xl) ≥ dM (xi, xj) + δijkl
}
.
⇒ This function is non-convex
5 / 14
Proposed Method:Algorithm
Singular value decomposition of M:
M = UΣU⊤
= · · · us · · ·
...
σs
...
· · · us · · ·
⊤
.
Define D:
D =
1
2
k∑
s=1
σ−1
s us u⊤
s .
, where k is number of singular values that smaller than ϵ.
Transform into this convex optimization by using D.
min
M ∈Sd
+
∑
(i,j,k,l)∈A
[
ξ(i,j,k,l) +
⟨
M, xij x⊤
ij − xkl x⊤
kl
⟩]
+
γ
2
Tr(M⊤
DM)
, where D is fixed.
6 / 14
Proposed Method:Algorithm
Proximal Gradient Descent
Key Points
They prove the convergence of our optimization algorithm.
k is hyper-parameter.
▶ ϵ is adaptively determined.
7 / 14
Experiment:Synthetic Data
Data:
1. Make T ∈ Sd
+, where rank(T) is e.
2. Quadrupletwise constraints that are satisfied on Mahalanobis Distance of T, split for
train data A, validation data V, test data T.
Setting:
▶ d = 100
▶ e = 10
▶ |A| = |V| = |T | = 104
▶ γ is tuned in the range of
{
10−2,10−2,1,10,102
}
▶ k is tuned from 5 to 20
Compared Methods:
▶ ML: No regularization
▶ ML+Trace: Trace Norm regularization
▶ ML+Fantope: Fantope regularization
▶ ML+capped: Proposed Method
8 / 14
Experiment:Synthetic Data
Accuracy and rank(M).
Method Accuracy rank(M)
ML 85.62% 53
ML + Trace 88.44% 41
ML + Fantope 95.50% 10
ML + capped 95.43% 10
Table 1: Synthetic experiment results.
⇒ Fantope reg and Capped Norm reg are both better than other method.
9 / 14
Experiment:Synthetic Data
Accuracy on changing hyper-parameter k
Rank k
6 8 10 12 14 16 18 20
Accuracy%
88
89
90
91
92
93
94
95
96
ML+Trace
ML+Fantope
Our method
⇒ Proposed method mostly outperforms Fantope reg.
⇒ Propsoed method performs more stable than Fantope reg.
10 / 14
Experiment:Labeled Faces in The Wild
Task: Deciding if two face images are from the same person.
Data:
▶ 13233 images from 5749 persons.
▶ Use SIFT feature.
Setting:
▶ Use pairwise constraints.
▶ γ is tuned in
{
10−2,10−1,1,10,102
}
▶ k is tuned in {30,35,40,45,50,55,60,65,70}
Compared Methods:
▶ IDENTITY: Euclidean Distance
▶ MAHALANOBIS: Traditional Mahalanobis Distance
▶ KISSME: [3]
▶ ITML: [4]
▶ LDML: [5]
11 / 14
Experiment:Labeled Faces in The Wild
ROC curve
False Positive Rate (FPR)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
TruePositiveRate(TPR)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
ROC
CAP (0.817)
FANTOPE (0.814)
KISSME (0.806)
ITML (0.794)
LDML (0.797)
IDENTITY (0.675)
MAHAL (0.748)
⇒ Proposed method is better than others.
12 / 14
Experiment:Labeled Faces in The Wild
Accuracy on changing hyper-parameter k.
Rank k
30 35 40 45 50 55 60 65 70
Accuracy%
79.5
80
80.5
81
81.5
82
82.5
ML+Trace
ML+Fantope
Our method
⇒ Proposed method get better results than metric learning with Fantope
regularization.
13 / 14
Conclusion
They proposed a novel low-rank regularization, Capped Trace Norm regularization.
Proposed algorithm for optimization problem and prove convergence of that.
Experimental results show that our method outperforms the state-of-the-art
metric learning methods.
14 / 14
Reference I
Zhouyuan Huo, Feiping Nie, and Heng Huang.
Robust and effective metric learning using capped trace norm: Metric learning via
capped trace norm.
In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge
Discovery and Data Mining, pp. 1605–1614. ACM, 2016.
Aur´elien Bellet, Amaury Habrard, and Marc Sebban.
A survey on metric learning for feature vectors and structured data.
arXiv preprint arXiv:1306.6709, 2013.
Martin Koestinger, Martin Hirzer, Paul Wohlhart, Peter M Roth, and Horst
Bischof.
Large scale metric learning from equivalence constraints.
In 2012 IEEE Conference on Computer Vision and Pattern Recognition, pp.
2288–2295. IEEE, 2012.
14 / 14
Reference II
Jason V Davis, Brian Kulis, Prateek Jain, Suvrit Sra, and Inderjit S Dhillon.
Information-theoretic metric learning.
In Proceedings of the 24th international conference on Machine learning, pp.
209–216. ACM, 2007.
Matthieu Guillaumin, Jakob Verbeek, and Cordelia Schmid.
Is that you? metric learning approaches for face identification.
In 2009 IEEE 12th International Conference on Computer Vision, pp. 498–505.
IEEE, 2009.
14 / 14

More Related Content

What's hot

Simplified Runtime Analysis of Estimation of Distribution Algorithms
Simplified Runtime Analysis of Estimation of Distribution AlgorithmsSimplified Runtime Analysis of Estimation of Distribution Algorithms
Simplified Runtime Analysis of Estimation of Distribution AlgorithmsPK Lehre
 
Estimating Future Initial Margin with Machine Learning
Estimating Future Initial Margin with Machine LearningEstimating Future Initial Margin with Machine Learning
Estimating Future Initial Margin with Machine LearningAndres Hernandez
 
MLHEP 2015: Introductory Lecture #3
MLHEP 2015: Introductory Lecture #3MLHEP 2015: Introductory Lecture #3
MLHEP 2015: Introductory Lecture #3arogozhnikov
 
Continuous and Discrete-Time Analysis of SGD
Continuous and Discrete-Time Analysis of SGDContinuous and Discrete-Time Analysis of SGD
Continuous and Discrete-Time Analysis of SGDValentin De Bortoli
 
MLHEP 2015: Introductory Lecture #4
MLHEP 2015: Introductory Lecture #4MLHEP 2015: Introductory Lecture #4
MLHEP 2015: Introductory Lecture #4arogozhnikov
 
From L to N: Nonlinear Predictors in Generalized Models
From L to N: Nonlinear Predictors in Generalized ModelsFrom L to N: Nonlinear Predictors in Generalized Models
From L to N: Nonlinear Predictors in Generalized Modelshtstatistics
 
MLHEP Lectures - day 1, basic track
MLHEP Lectures - day 1, basic trackMLHEP Lectures - day 1, basic track
MLHEP Lectures - day 1, basic trackarogozhnikov
 
Multiplicative Interaction Models in R
Multiplicative Interaction Models in RMultiplicative Interaction Models in R
Multiplicative Interaction Models in Rhtstatistics
 
Coordinate sampler: A non-reversible Gibbs-like sampler
Coordinate sampler: A non-reversible Gibbs-like samplerCoordinate sampler: A non-reversible Gibbs-like sampler
Coordinate sampler: A non-reversible Gibbs-like samplerChristian Robert
 
MLHEP Lectures - day 2, basic track
MLHEP Lectures - day 2, basic trackMLHEP Lectures - day 2, basic track
MLHEP Lectures - day 2, basic trackarogozhnikov
 
Approximate Bayesian Computation with Quasi-Likelihoods
Approximate Bayesian Computation with Quasi-LikelihoodsApproximate Bayesian Computation with Quasi-Likelihoods
Approximate Bayesian Computation with Quasi-LikelihoodsStefano Cabras
 
MLHEP 2015: Introductory Lecture #2
MLHEP 2015: Introductory Lecture #2MLHEP 2015: Introductory Lecture #2
MLHEP 2015: Introductory Lecture #2arogozhnikov
 
Bayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear modelsBayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear modelsCaleb (Shiqiang) Jin
 
short course at CIRM, Bayesian Masterclass, October 2018
short course at CIRM, Bayesian Masterclass, October 2018short course at CIRM, Bayesian Masterclass, October 2018
short course at CIRM, Bayesian Masterclass, October 2018Christian Robert
 
Kriging and spatial design accelerated by orders of magnitude
Kriging and spatial design accelerated by orders of magnitudeKriging and spatial design accelerated by orders of magnitude
Kriging and spatial design accelerated by orders of magnitudeAlexander Litvinenko
 

What's hot (19)

Simplified Runtime Analysis of Estimation of Distribution Algorithms
Simplified Runtime Analysis of Estimation of Distribution AlgorithmsSimplified Runtime Analysis of Estimation of Distribution Algorithms
Simplified Runtime Analysis of Estimation of Distribution Algorithms
 
Estimating Future Initial Margin with Machine Learning
Estimating Future Initial Margin with Machine LearningEstimating Future Initial Margin with Machine Learning
Estimating Future Initial Margin with Machine Learning
 
MLHEP 2015: Introductory Lecture #3
MLHEP 2015: Introductory Lecture #3MLHEP 2015: Introductory Lecture #3
MLHEP 2015: Introductory Lecture #3
 
Continuous and Discrete-Time Analysis of SGD
Continuous and Discrete-Time Analysis of SGDContinuous and Discrete-Time Analysis of SGD
Continuous and Discrete-Time Analysis of SGD
 
MLHEP 2015: Introductory Lecture #4
MLHEP 2015: Introductory Lecture #4MLHEP 2015: Introductory Lecture #4
MLHEP 2015: Introductory Lecture #4
 
From L to N: Nonlinear Predictors in Generalized Models
From L to N: Nonlinear Predictors in Generalized ModelsFrom L to N: Nonlinear Predictors in Generalized Models
From L to N: Nonlinear Predictors in Generalized Models
 
MLHEP Lectures - day 1, basic track
MLHEP Lectures - day 1, basic trackMLHEP Lectures - day 1, basic track
MLHEP Lectures - day 1, basic track
 
Multiplicative Interaction Models in R
Multiplicative Interaction Models in RMultiplicative Interaction Models in R
Multiplicative Interaction Models in R
 
Coordinate sampler: A non-reversible Gibbs-like sampler
Coordinate sampler: A non-reversible Gibbs-like samplerCoordinate sampler: A non-reversible Gibbs-like sampler
Coordinate sampler: A non-reversible Gibbs-like sampler
 
MLHEP Lectures - day 2, basic track
MLHEP Lectures - day 2, basic trackMLHEP Lectures - day 2, basic track
MLHEP Lectures - day 2, basic track
 
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
 
Approximate Bayesian Computation with Quasi-Likelihoods
Approximate Bayesian Computation with Quasi-LikelihoodsApproximate Bayesian Computation with Quasi-Likelihoods
Approximate Bayesian Computation with Quasi-Likelihoods
 
18.1 combining models
18.1 combining models18.1 combining models
18.1 combining models
 
MLHEP 2015: Introductory Lecture #2
MLHEP 2015: Introductory Lecture #2MLHEP 2015: Introductory Lecture #2
MLHEP 2015: Introductory Lecture #2
 
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
 
Bayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear modelsBayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear models
 
short course at CIRM, Bayesian Masterclass, October 2018
short course at CIRM, Bayesian Masterclass, October 2018short course at CIRM, Bayesian Masterclass, October 2018
short course at CIRM, Bayesian Masterclass, October 2018
 
Distributed ADMM
Distributed ADMMDistributed ADMM
Distributed ADMM
 
Kriging and spatial design accelerated by orders of magnitude
Kriging and spatial design accelerated by orders of magnitudeKriging and spatial design accelerated by orders of magnitude
Kriging and spatial design accelerated by orders of magnitude
 

Similar to Summery of Robust and Effective Metric Learning Using Capped Trace Norm

Unbiased Bayes for Big Data
Unbiased Bayes for Big DataUnbiased Bayes for Big Data
Unbiased Bayes for Big DataChristian Robert
 
Bayesian Deep Learning
Bayesian Deep LearningBayesian Deep Learning
Bayesian Deep LearningRayKim51
 
Performance of Matching Algorithmsfor Signal Approximation
Performance of Matching Algorithmsfor Signal ApproximationPerformance of Matching Algorithmsfor Signal Approximation
Performance of Matching Algorithmsfor Signal Approximationiosrjce
 
Introduction to Reinforcement Learning for Molecular Design
Introduction to Reinforcement Learning for Molecular Design Introduction to Reinforcement Learning for Molecular Design
Introduction to Reinforcement Learning for Molecular Design Dan Elton
 
Monte Carlo Berkeley.pptx
Monte Carlo Berkeley.pptxMonte Carlo Berkeley.pptx
Monte Carlo Berkeley.pptxHaibinSu2
 
MAPE regression, seminar @ QUT (Brisbane)
MAPE regression, seminar @ QUT (Brisbane)MAPE regression, seminar @ QUT (Brisbane)
MAPE regression, seminar @ QUT (Brisbane)Arnaud de Myttenaere
 
2014-mo444-practical-assignment-01-paulo_faria
2014-mo444-practical-assignment-01-paulo_faria2014-mo444-practical-assignment-01-paulo_faria
2014-mo444-practical-assignment-01-paulo_fariaPaulo Faria
 
Sampling based approximation of confidence intervals for functions of genetic...
Sampling based approximation of confidence intervals for functions of genetic...Sampling based approximation of confidence intervals for functions of genetic...
Sampling based approximation of confidence intervals for functions of genetic...prettygully
 
Monte Carlo Statistical Methods
Monte Carlo Statistical MethodsMonte Carlo Statistical Methods
Monte Carlo Statistical MethodsChristian Robert
 
Workshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinWorkshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinChristian Robert
 
On learning statistical mixtures maximizing the complete likelihood
On learning statistical mixtures maximizing the complete likelihoodOn learning statistical mixtures maximizing the complete likelihood
On learning statistical mixtures maximizing the complete likelihoodFrank Nielsen
 
New data structures and algorithms for \\post-processing large data sets and ...
New data structures and algorithms for \\post-processing large data sets and ...New data structures and algorithms for \\post-processing large data sets and ...
New data structures and algorithms for \\post-processing large data sets and ...Alexander Litvinenko
 

Similar to Summery of Robust and Effective Metric Learning Using Capped Trace Norm (20)

Unbiased Bayes for Big Data
Unbiased Bayes for Big DataUnbiased Bayes for Big Data
Unbiased Bayes for Big Data
 
Bayesian Deep Learning
Bayesian Deep LearningBayesian Deep Learning
Bayesian Deep Learning
 
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
Program on Mathematical and Statistical Methods for Climate and the Earth Sys...
 
Performance of Matching Algorithmsfor Signal Approximation
Performance of Matching Algorithmsfor Signal ApproximationPerformance of Matching Algorithmsfor Signal Approximation
Performance of Matching Algorithmsfor Signal Approximation
 
Introduction to Reinforcement Learning for Molecular Design
Introduction to Reinforcement Learning for Molecular Design Introduction to Reinforcement Learning for Molecular Design
Introduction to Reinforcement Learning for Molecular Design
 
Monte Carlo Berkeley.pptx
Monte Carlo Berkeley.pptxMonte Carlo Berkeley.pptx
Monte Carlo Berkeley.pptx
 
Cmb part3
Cmb part3Cmb part3
Cmb part3
 
MAPE regression, seminar @ QUT (Brisbane)
MAPE regression, seminar @ QUT (Brisbane)MAPE regression, seminar @ QUT (Brisbane)
MAPE regression, seminar @ QUT (Brisbane)
 
2014-mo444-practical-assignment-01-paulo_faria
2014-mo444-practical-assignment-01-paulo_faria2014-mo444-practical-assignment-01-paulo_faria
2014-mo444-practical-assignment-01-paulo_faria
 
the ABC of ABC
the ABC of ABCthe ABC of ABC
the ABC of ABC
 
Sampling based approximation of confidence intervals for functions of genetic...
Sampling based approximation of confidence intervals for functions of genetic...Sampling based approximation of confidence intervals for functions of genetic...
Sampling based approximation of confidence intervals for functions of genetic...
 
Monte Carlo Statistical Methods
Monte Carlo Statistical MethodsMonte Carlo Statistical Methods
Monte Carlo Statistical Methods
 
MUMS: Bayesian, Fiducial, and Frequentist Conference - Model Selection in the...
MUMS: Bayesian, Fiducial, and Frequentist Conference - Model Selection in the...MUMS: Bayesian, Fiducial, and Frequentist Conference - Model Selection in the...
MUMS: Bayesian, Fiducial, and Frequentist Conference - Model Selection in the...
 
QMC: Operator Splitting Workshop, Incremental Learning-to-Learn with Statisti...
QMC: Operator Splitting Workshop, Incremental Learning-to-Learn with Statisti...QMC: Operator Splitting Workshop, Incremental Learning-to-Learn with Statisti...
QMC: Operator Splitting Workshop, Incremental Learning-to-Learn with Statisti...
 
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
Program on Quasi-Monte Carlo and High-Dimensional Sampling Methods for Applie...
 
Workshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael MartinWorkshop in honour of Don Poskitt and Gael Martin
Workshop in honour of Don Poskitt and Gael Martin
 
On learning statistical mixtures maximizing the complete likelihood
On learning statistical mixtures maximizing the complete likelihoodOn learning statistical mixtures maximizing the complete likelihood
On learning statistical mixtures maximizing the complete likelihood
 
Gtti 10032021
Gtti 10032021Gtti 10032021
Gtti 10032021
 
New data structures and algorithms for \\post-processing large data sets and ...
New data structures and algorithms for \\post-processing large data sets and ...New data structures and algorithms for \\post-processing large data sets and ...
New data structures and algorithms for \\post-processing large data sets and ...
 
Statistics chm 235
Statistics chm 235Statistics chm 235
Statistics chm 235
 

Recently uploaded

Dr. E. Muralinath_ Blood indices_clinical aspects
Dr. E. Muralinath_ Blood indices_clinical  aspectsDr. E. Muralinath_ Blood indices_clinical  aspects
Dr. E. Muralinath_ Blood indices_clinical aspectsmuralinath2
 
Climate Change Impacts on Terrestrial and Aquatic Ecosystems.pptx
Climate Change Impacts on Terrestrial and Aquatic Ecosystems.pptxClimate Change Impacts on Terrestrial and Aquatic Ecosystems.pptx
Climate Change Impacts on Terrestrial and Aquatic Ecosystems.pptxDiariAli
 
Taphonomy and Quality of the Fossil Record
Taphonomy and Quality of the  Fossil RecordTaphonomy and Quality of the  Fossil Record
Taphonomy and Quality of the Fossil RecordSangram Sahoo
 
Lipids: types, structure and important functions.
Lipids: types, structure and important functions.Lipids: types, structure and important functions.
Lipids: types, structure and important functions.Cherry
 
CYTOGENETIC MAP................ ppt.pptx
CYTOGENETIC MAP................ ppt.pptxCYTOGENETIC MAP................ ppt.pptx
CYTOGENETIC MAP................ ppt.pptxCherry
 
ONLINE VOTING SYSTEM SE Project for vote
ONLINE VOTING SYSTEM SE Project for voteONLINE VOTING SYSTEM SE Project for vote
ONLINE VOTING SYSTEM SE Project for voteRaunakRastogi4
 
TransientOffsetin14CAftertheCarringtonEventRecordedbyPolarTreeRings
TransientOffsetin14CAftertheCarringtonEventRecordedbyPolarTreeRingsTransientOffsetin14CAftertheCarringtonEventRecordedbyPolarTreeRings
TransientOffsetin14CAftertheCarringtonEventRecordedbyPolarTreeRingsSérgio Sacani
 
GBSN - Microbiology (Unit 5) Concept of isolation
GBSN - Microbiology (Unit 5) Concept of isolationGBSN - Microbiology (Unit 5) Concept of isolation
GBSN - Microbiology (Unit 5) Concept of isolationAreesha Ahmad
 
GBSN - Microbiology (Unit 3)Defense Mechanism of the body
GBSN - Microbiology (Unit 3)Defense Mechanism of the body GBSN - Microbiology (Unit 3)Defense Mechanism of the body
GBSN - Microbiology (Unit 3)Defense Mechanism of the body Areesha Ahmad
 
module for grade 9 for distance learning
module for grade 9 for distance learningmodule for grade 9 for distance learning
module for grade 9 for distance learninglevieagacer
 
Porella : features, morphology, anatomy, reproduction etc.
Porella : features, morphology, anatomy, reproduction etc.Porella : features, morphology, anatomy, reproduction etc.
Porella : features, morphology, anatomy, reproduction etc.Cherry
 
Efficient spin-up of Earth System Models usingsequence acceleration
Efficient spin-up of Earth System Models usingsequence accelerationEfficient spin-up of Earth System Models usingsequence acceleration
Efficient spin-up of Earth System Models usingsequence accelerationSérgio Sacani
 
FAIRSpectra - Enabling the FAIRification of Spectroscopy and Spectrometry
FAIRSpectra - Enabling the FAIRification of Spectroscopy and SpectrometryFAIRSpectra - Enabling the FAIRification of Spectroscopy and Spectrometry
FAIRSpectra - Enabling the FAIRification of Spectroscopy and SpectrometryAlex Henderson
 
FS P2 COMBO MSTA LAST PUSH past exam papers.
FS P2 COMBO MSTA LAST PUSH past exam papers.FS P2 COMBO MSTA LAST PUSH past exam papers.
FS P2 COMBO MSTA LAST PUSH past exam papers.takadzanijustinmaime
 
FAIRSpectra - Enabling the FAIRification of Analytical Science
FAIRSpectra - Enabling the FAIRification of Analytical ScienceFAIRSpectra - Enabling the FAIRification of Analytical Science
FAIRSpectra - Enabling the FAIRification of Analytical ScienceAlex Henderson
 
(May 9, 2024) Enhanced Ultrafast Vector Flow Imaging (VFI) Using Multi-Angle ...
(May 9, 2024) Enhanced Ultrafast Vector Flow Imaging (VFI) Using Multi-Angle ...(May 9, 2024) Enhanced Ultrafast Vector Flow Imaging (VFI) Using Multi-Angle ...
(May 9, 2024) Enhanced Ultrafast Vector Flow Imaging (VFI) Using Multi-Angle ...Scintica Instrumentation
 
Genome sequencing,shotgun sequencing.pptx
Genome sequencing,shotgun sequencing.pptxGenome sequencing,shotgun sequencing.pptx
Genome sequencing,shotgun sequencing.pptxCherry
 
development of diagnostic enzyme assay to detect leuser virus
development of diagnostic enzyme assay to detect leuser virusdevelopment of diagnostic enzyme assay to detect leuser virus
development of diagnostic enzyme assay to detect leuser virusNazaninKarimi6
 
Cot curve, melting temperature, unique and repetitive DNA
Cot curve, melting temperature, unique and repetitive DNACot curve, melting temperature, unique and repetitive DNA
Cot curve, melting temperature, unique and repetitive DNACherry
 
COMPOSTING : types of compost, merits and demerits
COMPOSTING : types of compost, merits and demeritsCOMPOSTING : types of compost, merits and demerits
COMPOSTING : types of compost, merits and demeritsCherry
 

Recently uploaded (20)

Dr. E. Muralinath_ Blood indices_clinical aspects
Dr. E. Muralinath_ Blood indices_clinical  aspectsDr. E. Muralinath_ Blood indices_clinical  aspects
Dr. E. Muralinath_ Blood indices_clinical aspects
 
Climate Change Impacts on Terrestrial and Aquatic Ecosystems.pptx
Climate Change Impacts on Terrestrial and Aquatic Ecosystems.pptxClimate Change Impacts on Terrestrial and Aquatic Ecosystems.pptx
Climate Change Impacts on Terrestrial and Aquatic Ecosystems.pptx
 
Taphonomy and Quality of the Fossil Record
Taphonomy and Quality of the  Fossil RecordTaphonomy and Quality of the  Fossil Record
Taphonomy and Quality of the Fossil Record
 
Lipids: types, structure and important functions.
Lipids: types, structure and important functions.Lipids: types, structure and important functions.
Lipids: types, structure and important functions.
 
CYTOGENETIC MAP................ ppt.pptx
CYTOGENETIC MAP................ ppt.pptxCYTOGENETIC MAP................ ppt.pptx
CYTOGENETIC MAP................ ppt.pptx
 
ONLINE VOTING SYSTEM SE Project for vote
ONLINE VOTING SYSTEM SE Project for voteONLINE VOTING SYSTEM SE Project for vote
ONLINE VOTING SYSTEM SE Project for vote
 
TransientOffsetin14CAftertheCarringtonEventRecordedbyPolarTreeRings
TransientOffsetin14CAftertheCarringtonEventRecordedbyPolarTreeRingsTransientOffsetin14CAftertheCarringtonEventRecordedbyPolarTreeRings
TransientOffsetin14CAftertheCarringtonEventRecordedbyPolarTreeRings
 
GBSN - Microbiology (Unit 5) Concept of isolation
GBSN - Microbiology (Unit 5) Concept of isolationGBSN - Microbiology (Unit 5) Concept of isolation
GBSN - Microbiology (Unit 5) Concept of isolation
 
GBSN - Microbiology (Unit 3)Defense Mechanism of the body
GBSN - Microbiology (Unit 3)Defense Mechanism of the body GBSN - Microbiology (Unit 3)Defense Mechanism of the body
GBSN - Microbiology (Unit 3)Defense Mechanism of the body
 
module for grade 9 for distance learning
module for grade 9 for distance learningmodule for grade 9 for distance learning
module for grade 9 for distance learning
 
Porella : features, morphology, anatomy, reproduction etc.
Porella : features, morphology, anatomy, reproduction etc.Porella : features, morphology, anatomy, reproduction etc.
Porella : features, morphology, anatomy, reproduction etc.
 
Efficient spin-up of Earth System Models usingsequence acceleration
Efficient spin-up of Earth System Models usingsequence accelerationEfficient spin-up of Earth System Models usingsequence acceleration
Efficient spin-up of Earth System Models usingsequence acceleration
 
FAIRSpectra - Enabling the FAIRification of Spectroscopy and Spectrometry
FAIRSpectra - Enabling the FAIRification of Spectroscopy and SpectrometryFAIRSpectra - Enabling the FAIRification of Spectroscopy and Spectrometry
FAIRSpectra - Enabling the FAIRification of Spectroscopy and Spectrometry
 
FS P2 COMBO MSTA LAST PUSH past exam papers.
FS P2 COMBO MSTA LAST PUSH past exam papers.FS P2 COMBO MSTA LAST PUSH past exam papers.
FS P2 COMBO MSTA LAST PUSH past exam papers.
 
FAIRSpectra - Enabling the FAIRification of Analytical Science
FAIRSpectra - Enabling the FAIRification of Analytical ScienceFAIRSpectra - Enabling the FAIRification of Analytical Science
FAIRSpectra - Enabling the FAIRification of Analytical Science
 
(May 9, 2024) Enhanced Ultrafast Vector Flow Imaging (VFI) Using Multi-Angle ...
(May 9, 2024) Enhanced Ultrafast Vector Flow Imaging (VFI) Using Multi-Angle ...(May 9, 2024) Enhanced Ultrafast Vector Flow Imaging (VFI) Using Multi-Angle ...
(May 9, 2024) Enhanced Ultrafast Vector Flow Imaging (VFI) Using Multi-Angle ...
 
Genome sequencing,shotgun sequencing.pptx
Genome sequencing,shotgun sequencing.pptxGenome sequencing,shotgun sequencing.pptx
Genome sequencing,shotgun sequencing.pptx
 
development of diagnostic enzyme assay to detect leuser virus
development of diagnostic enzyme assay to detect leuser virusdevelopment of diagnostic enzyme assay to detect leuser virus
development of diagnostic enzyme assay to detect leuser virus
 
Cot curve, melting temperature, unique and repetitive DNA
Cot curve, melting temperature, unique and repetitive DNACot curve, melting temperature, unique and repetitive DNA
Cot curve, melting temperature, unique and repetitive DNA
 
COMPOSTING : types of compost, merits and demerits
COMPOSTING : types of compost, merits and demeritsCOMPOSTING : types of compost, merits and demerits
COMPOSTING : types of compost, merits and demerits
 

Summery of Robust and Effective Metric Learning Using Capped Trace Norm

  • 1. [Summery of] Robust and Effective Metric Learning Using Capped Trace Norm [1] (KDD 2016) shiba44
  • 3. Introduction Metric Learning aims automatically learning from train data. similar dissimilar Example: Learning distance of image They introduce low-rank regularization for Mahalanobis distance metric learning. dM (xi, xj) √ (xi − xj)⊤M(xi − xj) 1 / 14
  • 4. Background:Weak Constraints Pairwise similar dissimilar Tripletwise similar dissimilar Quadrupletwise similar dissimilar Pairwise: dM (xi, xj) is smaller, dM (xi′, xj′) is bigger. Tripletwise: dM (xi, xj) is smaller than dM (xj, xk). Quadrupletwise: dM (xi, xj) is smaller than dM (xk, xl). 2 / 14
  • 5. Background:low-rank and existing method M is low-rank → Mahalanobis distance is defined in low dim space. Rank minimization is NP Hard [2] ⇒ Use approximation of that. Trace Norm regularization ▶ Minimizes sum of all singular values σ(M). Reg(M) ∑ s σs(M). Changing of one large singular value affects overall. Fantope regularization ▶ Minimizes sum of k smallest singular values. Reg(M) k∑ s σs(M). Be sensitive for hyper-parameter k. 3 / 14
  • 6. Proposed Method In proposed method, use Capped Trace Norm regularization. Capped Trace Norm regularization Only minimizes the singular values that are smaller than ϵ. Reg(M) ∑ s min(σs(M),ϵ). Reduce the effect of changing large singular value. Stable for hyper-parameter than Fantope. 4 / 14
  • 7. Proposed Method Optimization Problem: min M ∈Sd + ∑ (i,j,k,l)∈A [ δijkl + ⟨ M, xij x⊤ ij − xkl x⊤ kl ⟩] + Degree of violation for quadrupletwise constraints + γ 2 ∑ s min(σs(M),ϵ) Regularization term ,where A = { (i, j, k,l) : dM (xk, xl) ≥ dM (xi, xj) + δijkl } . ⇒ This function is non-convex 5 / 14
  • 8. Proposed Method:Algorithm Singular value decomposition of M: M = UΣU⊤ = · · · us · · · ... σs ... · · · us · · · ⊤ . Define D: D = 1 2 k∑ s=1 σ−1 s us u⊤ s . , where k is number of singular values that smaller than ϵ. Transform into this convex optimization by using D. min M ∈Sd + ∑ (i,j,k,l)∈A [ ξ(i,j,k,l) + ⟨ M, xij x⊤ ij − xkl x⊤ kl ⟩] + γ 2 Tr(M⊤ DM) , where D is fixed. 6 / 14
  • 9. Proposed Method:Algorithm Proximal Gradient Descent Key Points They prove the convergence of our optimization algorithm. k is hyper-parameter. ▶ ϵ is adaptively determined. 7 / 14
  • 10. Experiment:Synthetic Data Data: 1. Make T ∈ Sd +, where rank(T) is e. 2. Quadrupletwise constraints that are satisfied on Mahalanobis Distance of T, split for train data A, validation data V, test data T. Setting: ▶ d = 100 ▶ e = 10 ▶ |A| = |V| = |T | = 104 ▶ γ is tuned in the range of { 10−2,10−2,1,10,102 } ▶ k is tuned from 5 to 20 Compared Methods: ▶ ML: No regularization ▶ ML+Trace: Trace Norm regularization ▶ ML+Fantope: Fantope regularization ▶ ML+capped: Proposed Method 8 / 14
  • 11. Experiment:Synthetic Data Accuracy and rank(M). Method Accuracy rank(M) ML 85.62% 53 ML + Trace 88.44% 41 ML + Fantope 95.50% 10 ML + capped 95.43% 10 Table 1: Synthetic experiment results. ⇒ Fantope reg and Capped Norm reg are both better than other method. 9 / 14
  • 12. Experiment:Synthetic Data Accuracy on changing hyper-parameter k Rank k 6 8 10 12 14 16 18 20 Accuracy% 88 89 90 91 92 93 94 95 96 ML+Trace ML+Fantope Our method ⇒ Proposed method mostly outperforms Fantope reg. ⇒ Propsoed method performs more stable than Fantope reg. 10 / 14
  • 13. Experiment:Labeled Faces in The Wild Task: Deciding if two face images are from the same person. Data: ▶ 13233 images from 5749 persons. ▶ Use SIFT feature. Setting: ▶ Use pairwise constraints. ▶ γ is tuned in { 10−2,10−1,1,10,102 } ▶ k is tuned in {30,35,40,45,50,55,60,65,70} Compared Methods: ▶ IDENTITY: Euclidean Distance ▶ MAHALANOBIS: Traditional Mahalanobis Distance ▶ KISSME: [3] ▶ ITML: [4] ▶ LDML: [5] 11 / 14
  • 14. Experiment:Labeled Faces in The Wild ROC curve False Positive Rate (FPR) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 TruePositiveRate(TPR) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 ROC CAP (0.817) FANTOPE (0.814) KISSME (0.806) ITML (0.794) LDML (0.797) IDENTITY (0.675) MAHAL (0.748) ⇒ Proposed method is better than others. 12 / 14
  • 15. Experiment:Labeled Faces in The Wild Accuracy on changing hyper-parameter k. Rank k 30 35 40 45 50 55 60 65 70 Accuracy% 79.5 80 80.5 81 81.5 82 82.5 ML+Trace ML+Fantope Our method ⇒ Proposed method get better results than metric learning with Fantope regularization. 13 / 14
  • 16. Conclusion They proposed a novel low-rank regularization, Capped Trace Norm regularization. Proposed algorithm for optimization problem and prove convergence of that. Experimental results show that our method outperforms the state-of-the-art metric learning methods. 14 / 14
  • 17. Reference I Zhouyuan Huo, Feiping Nie, and Heng Huang. Robust and effective metric learning using capped trace norm: Metric learning via capped trace norm. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 1605–1614. ACM, 2016. Aur´elien Bellet, Amaury Habrard, and Marc Sebban. A survey on metric learning for feature vectors and structured data. arXiv preprint arXiv:1306.6709, 2013. Martin Koestinger, Martin Hirzer, Paul Wohlhart, Peter M Roth, and Horst Bischof. Large scale metric learning from equivalence constraints. In 2012 IEEE Conference on Computer Vision and Pattern Recognition, pp. 2288–2295. IEEE, 2012. 14 / 14
  • 18. Reference II Jason V Davis, Brian Kulis, Prateek Jain, Suvrit Sra, and Inderjit S Dhillon. Information-theoretic metric learning. In Proceedings of the 24th international conference on Machine learning, pp. 209–216. ACM, 2007. Matthieu Guillaumin, Jakob Verbeek, and Cordelia Schmid. Is that you? metric learning approaches for face identification. In 2009 IEEE 12th International Conference on Computer Vision, pp. 498–505. IEEE, 2009. 14 / 14