SlideShare a Scribd company logo
1 of 18
Download to read offline
CVPR reading
2015.7.24
Akisato Kimura
<@_akisato, akisato@ieee.org>
1
Paper to read
CVPR2015 (poster)
2
1-page summary
• A method for refining a pre-trained random forest
– Comparable to RF with much more nodes of decision trees
– Better than RF with the same size of decision trees
3
Random forest
• An ensemble of decision trees trained by bootstrap
sampling & random feature selection
Decision tree
𝒙𝒙𝑖𝑖
�𝑦𝑦𝑖𝑖 = 𝑓𝑓𝑡𝑡(𝒙𝒙𝑖𝑖)
(returns a MAP prediction) 4
Random forest
• An ensemble of decision trees trained by bootstrap
sampling & random feature selection
𝒙𝒙𝑖𝑖
�𝑦𝑦𝑖𝑖 = 𝑓𝑓1(𝒙𝒙𝑖𝑖)
(returns a MAP prediction)
𝒙𝒙𝑖𝑖
�𝑦𝑦𝑖𝑖 = 𝑓𝑓2(𝒙𝒙𝑖𝑖)
𝒙𝒙𝑖𝑖
�𝑦𝑦𝑖𝑖 = 𝑓𝑓𝑇𝑇(𝒙𝒙𝑖𝑖)
Random forest
5
Reformulation of a decision tree
• A prediction 𝑓𝑓 can be divided into 2 components
Decision tree
𝒙𝒙𝑖𝑖
�𝑦𝑦𝑖𝑖 = 𝑓𝑓𝑡𝑡(𝒙𝒙𝑖𝑖)
(returns a MAP prediction)
𝜙𝜙𝑡𝑡 𝒙𝒙𝑖𝑖 = 0, 0, 1, 0 𝑇𝑇
Indicating which path to a leaf node is selected,
can be represented by a binary vector.
𝑤𝑤𝑡𝑡(𝑦𝑦) = 0.2, 0.5, 0.8, 0.1 𝑇𝑇
Storing a posterior probability of 𝑦𝑦 at each leaf node,
can be represented by a real vector.
𝑓𝑓𝑡𝑡 𝒙𝒙𝑖𝑖 = arg max
𝑦𝑦
𝑤𝑤𝑡𝑡(𝑦𝑦) ⋅ 𝜙𝜙𝑡𝑡(𝒙𝒙𝑖𝑖)
Indicator vector
Leaf vector
6
Reformulation of a random forest
• A prediction 𝑓𝑓 can be divided into 2 components
𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖
0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖
0.2 0.5 0.8 0.1 0.3 0.7 0.1 0.2 0.1 0.1 0.5 0.3𝑊𝑊 𝑦𝑦
Random forest 𝑓𝑓 𝒙𝒙𝑖𝑖 = arg max
𝑦𝑦
𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙𝑖𝑖)
7
Look like a SVM classifier
Global refinement
• Optimize a leaf vector (weights) 𝑊𝑊(𝑦𝑦),
while maintaining the indicator vector (structure) Φ(𝑥𝑥)
𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖
0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖
0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦
Random forest 𝑓𝑓 𝒙𝒙𝑖𝑖 = arg max
𝑦𝑦
𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙𝑖𝑖)
8
Global refinement
• Optimize a leaf vector (weights) 𝑊𝑊(𝑦𝑦),
while maintaining the indicator vector (structure) Φ(𝑥𝑥)
𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖
0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖
0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦
Random forest 𝑓𝑓 𝒙𝒙𝑖𝑖 = arg max
𝑦𝑦
𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙𝑖𝑖)
This optimization can be regarded as a linear classification problem,
where an indicator vector Φ(𝒙𝒙) is a new representation of a sample 𝒙𝒙.
[Note] In standard random forest, the trees are independently optimized.
This optimization effectively utilizes complementary information among trees.
9
Global refinement
• Optimize a leaf vector (weights) 𝑊𝑊(𝑦𝑦),
while maintaining the indicator vector (structure) Φ(𝑥𝑥)
𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖
0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖
0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦
Random forest 𝑓𝑓 𝒙𝒙𝑖𝑖 = arg max
𝑦𝑦
𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙𝑖𝑖)
This optimization can be regarded as a linear classification problem,
where an indicator vector Φ(𝒙𝒙) is a new representation of a sample 𝒙𝒙.
A sample Φ(𝑥𝑥) is highly sparse  Liblinear well suits this problem.
It can be easily extended to a regression problem.
10
Global pruning
• Adjacent leaves with nearly-zero weights 𝑊𝑊(𝑦𝑦)
do not contribute to the final result.  merge them.
𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖
0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖
0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦
Random forest 𝑓𝑓 𝒙𝒙 = arg max
𝑦𝑦
𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙)
0
0.1
0
0.1
11
Global pruning
• Adjacent leaves with nearly-zero weights 𝑊𝑊(𝑦𝑦)
do not contribute to the final result.  merge them.
𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖
0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖
0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦
Random forest 𝑓𝑓 𝒙𝒙 = arg max
𝑦𝑦
𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙)
0
0.1
0
0.1
1. Optimize leaf vectors 𝑊𝑊 𝑦𝑦 ∀𝑦𝑦
2. Prune a certain percentage of insignificant leaves
(significance = sum of elements in leaf vectors)
3. Update indicator vectors Φ(𝑥𝑥) for all the training samples
4. Repeat 1-3 until satisfying certain criterion, e.g.
a. the size of the random forest is smaller than predefined,
b. the prediction accuracy achieves best on a validation set.
12
Data sets for experiments
13
Experimental results
• ADF/ARF - alternating decision (regression) forest [Schulter+ ICCV13]
• Refined-A - Proposed method with the “accuracy” criterion
• Refined-E - Proposed method with “over-pruning”
(Accuracy is comparable to the original RF, but the size is much smaller.)
• Metrics - Error rate for classification, RMSE for regression.
• # trees = 100, max. depth = 10, 15 or 25 depending on the size of the training data.
• 60% for training, 40% for testing. 14
Parameter analysis
• The proposed method achieved better performances
than RFs with the same tree parameters (e.g. the number and
depth of trees)
15
(for MNIST data)
Parameter analysis
• The proposed method accelerates both training and
testing steps
16
(for MNIST data)
Number of dimensions used
on each node splitting
Number of samples used in
each decision tree
Best for RFBest for
proposed
Time for testingfast slow Time for trainingfast slow
Less sensitive More samples
needed
Applications
• Kinect body part classification
• Age regression from face images
17
Task-specific features 
Last words
• Simple, easy to implement, but effective
• Can be applicable to other classifiers
18

More Related Content

What's hot

Introduction to random forest and gradient boosting methods a lecture
Introduction to random forest and gradient boosting methods   a lectureIntroduction to random forest and gradient boosting methods   a lecture
Introduction to random forest and gradient boosting methods a lectureShreyas S K
 
Random Forest and KNN is fun
Random Forest and KNN is funRandom Forest and KNN is fun
Random Forest and KNN is funZhen Li
 
Understanding random forests
Understanding random forestsUnderstanding random forests
Understanding random forestsMarc Garcia
 
Machine learning basics using trees algorithm (Random forest, Gradient Boosting)
Machine learning basics using trees algorithm (Random forest, Gradient Boosting)Machine learning basics using trees algorithm (Random forest, Gradient Boosting)
Machine learning basics using trees algorithm (Random forest, Gradient Boosting)Parth Khare
 
Decision Trees for Classification: A Machine Learning Algorithm
Decision Trees for Classification: A Machine Learning AlgorithmDecision Trees for Classification: A Machine Learning Algorithm
Decision Trees for Classification: A Machine Learning AlgorithmPalin analytics
 
Decision Trees
Decision TreesDecision Trees
Decision TreesStudent
 
Lect 3 background mathematics
Lect 3 background mathematicsLect 3 background mathematics
Lect 3 background mathematicshktripathy
 
Machine Learning Feature Selection - Random Forest
Machine Learning Feature Selection - Random Forest Machine Learning Feature Selection - Random Forest
Machine Learning Feature Selection - Random Forest Rupak Roy
 
Data Compression in Data mining and Business Intelligencs
Data Compression in Data mining and Business Intelligencs Data Compression in Data mining and Business Intelligencs
Data Compression in Data mining and Business Intelligencs ShahDhruv21
 
Unsupervised Learning Techniques to Diversifying and Pruning Random Forest
Unsupervised Learning Techniques to Diversifying and Pruning Random ForestUnsupervised Learning Techniques to Diversifying and Pruning Random Forest
Unsupervised Learning Techniques to Diversifying and Pruning Random ForestMohamed Medhat Gaber
 
Random Forest Algorithm - Random Forest Explained | Random Forest In Machine ...
Random Forest Algorithm - Random Forest Explained | Random Forest In Machine ...Random Forest Algorithm - Random Forest Explained | Random Forest In Machine ...
Random Forest Algorithm - Random Forest Explained | Random Forest In Machine ...Simplilearn
 
Ml6 decision trees
Ml6 decision treesMl6 decision trees
Ml6 decision treesankit_ppt
 
Understanding Bagging and Boosting
Understanding Bagging and BoostingUnderstanding Bagging and Boosting
Understanding Bagging and BoostingMohit Rajput
 
Machine Learning with Decision trees
Machine Learning with Decision treesMachine Learning with Decision trees
Machine Learning with Decision treesKnoldus Inc.
 
Random Forest / Bootstrap Aggregation
Random Forest / Bootstrap AggregationRandom Forest / Bootstrap Aggregation
Random Forest / Bootstrap AggregationRupak Roy
 
Data Science - Part V - Decision Trees & Random Forests
Data Science - Part V - Decision Trees & Random Forests Data Science - Part V - Decision Trees & Random Forests
Data Science - Part V - Decision Trees & Random Forests Derek Kane
 
Understanding the Machine Learning Algorithms
Understanding the Machine Learning AlgorithmsUnderstanding the Machine Learning Algorithms
Understanding the Machine Learning AlgorithmsRupak Roy
 

What's hot (20)

Introduction to random forest and gradient boosting methods a lecture
Introduction to random forest and gradient boosting methods   a lectureIntroduction to random forest and gradient boosting methods   a lecture
Introduction to random forest and gradient boosting methods a lecture
 
Random Forest and KNN is fun
Random Forest and KNN is funRandom Forest and KNN is fun
Random Forest and KNN is fun
 
Understanding random forests
Understanding random forestsUnderstanding random forests
Understanding random forests
 
L4. Ensembles of Decision Trees
L4. Ensembles of Decision TreesL4. Ensembles of Decision Trees
L4. Ensembles of Decision Trees
 
Machine learning basics using trees algorithm (Random forest, Gradient Boosting)
Machine learning basics using trees algorithm (Random forest, Gradient Boosting)Machine learning basics using trees algorithm (Random forest, Gradient Boosting)
Machine learning basics using trees algorithm (Random forest, Gradient Boosting)
 
Decision Trees for Classification: A Machine Learning Algorithm
Decision Trees for Classification: A Machine Learning AlgorithmDecision Trees for Classification: A Machine Learning Algorithm
Decision Trees for Classification: A Machine Learning Algorithm
 
Decision Trees
Decision TreesDecision Trees
Decision Trees
 
Lect 3 background mathematics
Lect 3 background mathematicsLect 3 background mathematics
Lect 3 background mathematics
 
Random forest
Random forestRandom forest
Random forest
 
Ml7 bagging
Ml7 baggingMl7 bagging
Ml7 bagging
 
Machine Learning Feature Selection - Random Forest
Machine Learning Feature Selection - Random Forest Machine Learning Feature Selection - Random Forest
Machine Learning Feature Selection - Random Forest
 
Data Compression in Data mining and Business Intelligencs
Data Compression in Data mining and Business Intelligencs Data Compression in Data mining and Business Intelligencs
Data Compression in Data mining and Business Intelligencs
 
Unsupervised Learning Techniques to Diversifying and Pruning Random Forest
Unsupervised Learning Techniques to Diversifying and Pruning Random ForestUnsupervised Learning Techniques to Diversifying and Pruning Random Forest
Unsupervised Learning Techniques to Diversifying and Pruning Random Forest
 
Random Forest Algorithm - Random Forest Explained | Random Forest In Machine ...
Random Forest Algorithm - Random Forest Explained | Random Forest In Machine ...Random Forest Algorithm - Random Forest Explained | Random Forest In Machine ...
Random Forest Algorithm - Random Forest Explained | Random Forest In Machine ...
 
Ml6 decision trees
Ml6 decision treesMl6 decision trees
Ml6 decision trees
 
Understanding Bagging and Boosting
Understanding Bagging and BoostingUnderstanding Bagging and Boosting
Understanding Bagging and Boosting
 
Machine Learning with Decision trees
Machine Learning with Decision treesMachine Learning with Decision trees
Machine Learning with Decision trees
 
Random Forest / Bootstrap Aggregation
Random Forest / Bootstrap AggregationRandom Forest / Bootstrap Aggregation
Random Forest / Bootstrap Aggregation
 
Data Science - Part V - Decision Trees & Random Forests
Data Science - Part V - Decision Trees & Random Forests Data Science - Part V - Decision Trees & Random Forests
Data Science - Part V - Decision Trees & Random Forests
 
Understanding the Machine Learning Algorithms
Understanding the Machine Learning AlgorithmsUnderstanding the Machine Learning Algorithms
Understanding the Machine Learning Algorithms
 

Viewers also liked

Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in RUnderstanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in RManish Saraswat
 
Decision Forests and discriminant analysis
Decision Forests and discriminant analysisDecision Forests and discriminant analysis
Decision Forests and discriminant analysispotaters
 
random forests for ABC model choice and parameter estimation
random forests for ABC model choice and parameter estimationrandom forests for ABC model choice and parameter estimation
random forests for ABC model choice and parameter estimationChristian Robert
 
Bagging Decision Trees on Data Sets with Classification Noise
Bagging Decision Trees on Data Sets with Classification NoiseBagging Decision Trees on Data Sets with Classification Noise
Bagging Decision Trees on Data Sets with Classification NoiseNTNU
 
Understanding Random Forests: From Theory to Practice
Understanding Random Forests: From Theory to PracticeUnderstanding Random Forests: From Theory to Practice
Understanding Random Forests: From Theory to PracticeGilles Louppe
 
ICCV2015勉強会 顔関連論文のまとめ
ICCV2015勉強会 顔関連論文のまとめICCV2015勉強会 顔関連論文のまとめ
ICCV2015勉強会 顔関連論文のまとめKoichi Takahashi
 
直感的な単変量モデルでは予測できない「ワインの味」を多変量モデルで予測する
直感的な単変量モデルでは予測できない「ワインの味」を多変量モデルで予測する直感的な単変量モデルでは予測できない「ワインの味」を多変量モデルで予測する
直感的な単変量モデルでは予測できない「ワインの味」を多変量モデルで予測するTakashi J OZAKI
 
ランダムフォレスト
ランダムフォレストランダムフォレスト
ランダムフォレストKinki University
 
Feature Selection with R / in JP
Feature Selection with R / in JPFeature Selection with R / in JP
Feature Selection with R / in JPSercan Ahi
 
CVPR2015勉強会 Global Refinement of Random Forest
CVPR2015勉強会 Global Refinement of Random ForestCVPR2015勉強会 Global Refinement of Random Forest
CVPR2015勉強会 Global Refinement of Random ForestKoichi Takahashi
 
How to create great slides for presentations
How to create great slides for presentationsHow to create great slides for presentations
How to create great slides for presentationsmikejeffs
 

Viewers also liked (11)

Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in RUnderstanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
Understanding Deep Learning & Parameter Tuning with MXnet, H2o Package in R
 
Decision Forests and discriminant analysis
Decision Forests and discriminant analysisDecision Forests and discriminant analysis
Decision Forests and discriminant analysis
 
random forests for ABC model choice and parameter estimation
random forests for ABC model choice and parameter estimationrandom forests for ABC model choice and parameter estimation
random forests for ABC model choice and parameter estimation
 
Bagging Decision Trees on Data Sets with Classification Noise
Bagging Decision Trees on Data Sets with Classification NoiseBagging Decision Trees on Data Sets with Classification Noise
Bagging Decision Trees on Data Sets with Classification Noise
 
Understanding Random Forests: From Theory to Practice
Understanding Random Forests: From Theory to PracticeUnderstanding Random Forests: From Theory to Practice
Understanding Random Forests: From Theory to Practice
 
ICCV2015勉強会 顔関連論文のまとめ
ICCV2015勉強会 顔関連論文のまとめICCV2015勉強会 顔関連論文のまとめ
ICCV2015勉強会 顔関連論文のまとめ
 
直感的な単変量モデルでは予測できない「ワインの味」を多変量モデルで予測する
直感的な単変量モデルでは予測できない「ワインの味」を多変量モデルで予測する直感的な単変量モデルでは予測できない「ワインの味」を多変量モデルで予測する
直感的な単変量モデルでは予測できない「ワインの味」を多変量モデルで予測する
 
ランダムフォレスト
ランダムフォレストランダムフォレスト
ランダムフォレスト
 
Feature Selection with R / in JP
Feature Selection with R / in JPFeature Selection with R / in JP
Feature Selection with R / in JP
 
CVPR2015勉強会 Global Refinement of Random Forest
CVPR2015勉強会 Global Refinement of Random ForestCVPR2015勉強会 Global Refinement of Random Forest
CVPR2015勉強会 Global Refinement of Random Forest
 
How to create great slides for presentations
How to create great slides for presentationsHow to create great slides for presentations
How to create great slides for presentations
 

Similar to CVPR2015 reading "Global refinement of random forest"

MLHEP 2015: Introductory Lecture #2
MLHEP 2015: Introductory Lecture #2MLHEP 2015: Introductory Lecture #2
MLHEP 2015: Introductory Lecture #2arogozhnikov
 
RandomForests in artificial intelligence
RandomForests in artificial intelligenceRandomForests in artificial intelligence
RandomForests in artificial intelligencePriyadharshiniG41
 
Tree models with Scikit-Learn: Great models with little assumptions
Tree models with Scikit-Learn: Great models with little assumptionsTree models with Scikit-Learn: Great models with little assumptions
Tree models with Scikit-Learn: Great models with little assumptionsGilles Louppe
 
Tree net and_randomforests_2009
Tree net and_randomforests_2009Tree net and_randomforests_2009
Tree net and_randomforests_2009Matthew Magistrado
 
Machine learning in science and industry — day 2
Machine learning in science and industry — day 2Machine learning in science and industry — day 2
Machine learning in science and industry — day 2arogozhnikov
 
Elements of Statistical Learning 読み会 第2章
Elements of Statistical Learning 読み会 第2章Elements of Statistical Learning 読み会 第2章
Elements of Statistical Learning 読み会 第2章Tsuyoshi Sakama
 
DCWP_CVPR2023.pptx
DCWP_CVPR2023.pptxDCWP_CVPR2023.pptx
DCWP_CVPR2023.pptx건영 박
 
Molinier - Feature Selection for Tree Species Identification in Very High res...
Molinier - Feature Selection for Tree Species Identification in Very High res...Molinier - Feature Selection for Tree Species Identification in Very High res...
Molinier - Feature Selection for Tree Species Identification in Very High res...grssieee
 
Machine learning ppt and presentation code
Machine learning ppt and presentation codeMachine learning ppt and presentation code
Machine learning ppt and presentation codesharma239172
 
ABC short course: final chapters
ABC short course: final chaptersABC short course: final chapters
ABC short course: final chaptersChristian Robert
 
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920Karl Rudeen
 
Binary Vector Reconstruction via Discreteness-Aware Approximate Message Passing
Binary Vector Reconstruction via Discreteness-Aware Approximate Message PassingBinary Vector Reconstruction via Discreteness-Aware Approximate Message Passing
Binary Vector Reconstruction via Discreteness-Aware Approximate Message PassingRyo Hayakawa
 
Intro. to computational Physics ch2.pdf
Intro. to computational Physics ch2.pdfIntro. to computational Physics ch2.pdf
Intro. to computational Physics ch2.pdfJifarRaya
 

Similar to CVPR2015 reading "Global refinement of random forest" (20)

MLHEP 2015: Introductory Lecture #2
MLHEP 2015: Introductory Lecture #2MLHEP 2015: Introductory Lecture #2
MLHEP 2015: Introductory Lecture #2
 
RandomForests in artificial intelligence
RandomForests in artificial intelligenceRandomForests in artificial intelligence
RandomForests in artificial intelligence
 
Tree models with Scikit-Learn: Great models with little assumptions
Tree models with Scikit-Learn: Great models with little assumptionsTree models with Scikit-Learn: Great models with little assumptions
Tree models with Scikit-Learn: Great models with little assumptions
 
Curve fitting
Curve fittingCurve fitting
Curve fitting
 
Curve fitting
Curve fittingCurve fitting
Curve fitting
 
Tree net and_randomforests_2009
Tree net and_randomforests_2009Tree net and_randomforests_2009
Tree net and_randomforests_2009
 
Machine learning in science and industry — day 2
Machine learning in science and industry — day 2Machine learning in science and industry — day 2
Machine learning in science and industry — day 2
 
Elements of Statistical Learning 読み会 第2章
Elements of Statistical Learning 読み会 第2章Elements of Statistical Learning 読み会 第2章
Elements of Statistical Learning 読み会 第2章
 
ABC-Gibbs
ABC-GibbsABC-Gibbs
ABC-Gibbs
 
DCWP_CVPR2023.pptx
DCWP_CVPR2023.pptxDCWP_CVPR2023.pptx
DCWP_CVPR2023.pptx
 
Molinier - Feature Selection for Tree Species Identification in Very High res...
Molinier - Feature Selection for Tree Species Identification in Very High res...Molinier - Feature Selection for Tree Species Identification in Very High res...
Molinier - Feature Selection for Tree Species Identification in Very High res...
 
Machine learning ppt and presentation code
Machine learning ppt and presentation codeMachine learning ppt and presentation code
Machine learning ppt and presentation code
 
ABC short course: final chapters
ABC short course: final chaptersABC short course: final chapters
ABC short course: final chapters
 
the ABC of ABC
the ABC of ABCthe ABC of ABC
the ABC of ABC
 
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
fb69b412-97cb-4e8d-8a28-574c09557d35-160618025920
 
Project Paper
Project PaperProject Paper
Project Paper
 
Binary Vector Reconstruction via Discreteness-Aware Approximate Message Passing
Binary Vector Reconstruction via Discreteness-Aware Approximate Message PassingBinary Vector Reconstruction via Discreteness-Aware Approximate Message Passing
Binary Vector Reconstruction via Discreteness-Aware Approximate Message Passing
 
MSSISS riBART 20160321
MSSISS riBART 20160321MSSISS riBART 20160321
MSSISS riBART 20160321
 
Estimating Space-Time Covariance from Finite Sample Sets
Estimating Space-Time Covariance from Finite Sample SetsEstimating Space-Time Covariance from Finite Sample Sets
Estimating Space-Time Covariance from Finite Sample Sets
 
Intro. to computational Physics ch2.pdf
Intro. to computational Physics ch2.pdfIntro. to computational Physics ch2.pdf
Intro. to computational Physics ch2.pdf
 

More from Akisato Kimura

Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Akisato Kimura
 
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Akisato Kimura
 
多変量解析の一般化
多変量解析の一般化多変量解析の一般化
多変量解析の一般化Akisato Kimura
 
CVPR2016 reading - 特徴量学習とクロスモーダル転移について
CVPR2016 reading - 特徴量学習とクロスモーダル転移についてCVPR2016 reading - 特徴量学習とクロスモーダル転移について
CVPR2016 reading - 特徴量学習とクロスモーダル転移についてAkisato Kimura
 
NIPS2015 reading - Learning visual biases from human imagination
NIPS2015 reading - Learning visual biases from human imaginationNIPS2015 reading - Learning visual biases from human imagination
NIPS2015 reading - Learning visual biases from human imaginationAkisato Kimura
 
CVPR2015 reading "Understainding image virality" (in Japanese)
CVPR2015 reading "Understainding image virality" (in Japanese)CVPR2015 reading "Understainding image virality" (in Japanese)
CVPR2015 reading "Understainding image virality" (in Japanese)Akisato Kimura
 
Computational models of human visual attention driven by auditory cues
Computational models of human visual attention driven by auditory cuesComputational models of human visual attention driven by auditory cues
Computational models of human visual attention driven by auditory cuesAkisato Kimura
 
NIPS2014 reading - Top rank optimization in linear time
NIPS2014 reading - Top rank optimization in linear timeNIPS2014 reading - Top rank optimization in linear time
NIPS2014 reading - Top rank optimization in linear timeAkisato Kimura
 
CVPR2014 reading "Reconstructing storyline graphs for image recommendation fr...
CVPR2014 reading "Reconstructing storyline graphs for image recommendation fr...CVPR2014 reading "Reconstructing storyline graphs for image recommendation fr...
CVPR2014 reading "Reconstructing storyline graphs for image recommendation fr...Akisato Kimura
 
ICCV2013 reading: Learning to rank using privileged information
ICCV2013 reading: Learning to rank using privileged informationICCV2013 reading: Learning to rank using privileged information
ICCV2013 reading: Learning to rank using privileged informationAkisato Kimura
 
ACMMM 2013 reading: Large-scale visual sentiment ontology and detectors using...
ACMMM 2013 reading: Large-scale visual sentiment ontology and detectors using...ACMMM 2013 reading: Large-scale visual sentiment ontology and detectors using...
ACMMM 2013 reading: Large-scale visual sentiment ontology and detectors using...Akisato Kimura
 
IJCAI13 Paper review: Large-scale spectral clustering on graphs
IJCAI13 Paper review: Large-scale spectral clustering on graphsIJCAI13 Paper review: Large-scale spectral clustering on graphs
IJCAI13 Paper review: Large-scale spectral clustering on graphsAkisato Kimura
 
関西CVPR勉強会 2012.10.28
関西CVPR勉強会 2012.10.28関西CVPR勉強会 2012.10.28
関西CVPR勉強会 2012.10.28Akisato Kimura
 
関西CVPR勉強会 2012.7.29
関西CVPR勉強会 2012.7.29関西CVPR勉強会 2012.7.29
関西CVPR勉強会 2012.7.29Akisato Kimura
 
関西CVPRML勉強会 2012.2.18 (一般物体認識 - データセット)
関西CVPRML勉強会 2012.2.18 (一般物体認識 - データセット)関西CVPRML勉強会 2012.2.18 (一般物体認識 - データセット)
関西CVPRML勉強会 2012.2.18 (一般物体認識 - データセット)Akisato Kimura
 
関西CVPRML勉強会(特定物体認識) 2012.1.14
関西CVPRML勉強会(特定物体認識) 2012.1.14関西CVPRML勉強会(特定物体認識) 2012.1.14
関西CVPRML勉強会(特定物体認識) 2012.1.14Akisato Kimura
 
人間の視覚的注意を予測するモデル - 動的ベイジアンネットワークに基づく 最新のアプローチ -
人間の視覚的注意を予測するモデル - 動的ベイジアンネットワークに基づく 最新のアプローチ -人間の視覚的注意を予測するモデル - 動的ベイジアンネットワークに基づく 最新のアプローチ -
人間の視覚的注意を予測するモデル - 動的ベイジアンネットワークに基づく 最新のアプローチ -Akisato Kimura
 
IBIS2011 企画セッション「CV/PRで独自の進化を遂げる学習・最適化技術」 趣旨説明
IBIS2011 企画セッション「CV/PRで独自の進化を遂げる学習・最適化技術」 趣旨説明IBIS2011 企画セッション「CV/PRで独自の進化を遂げる学習・最適化技術」 趣旨説明
IBIS2011 企画セッション「CV/PRで独自の進化を遂げる学習・最適化技術」 趣旨説明Akisato Kimura
 
立命館大学 AMLコロキウム 2011.10.20
立命館大学 AMLコロキウム 2011.10.20立命館大学 AMLコロキウム 2011.10.20
立命館大学 AMLコロキウム 2011.10.20Akisato Kimura
 

More from Akisato Kimura (20)

Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
 
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
Paper reading - Dropout as a Bayesian Approximation: Representing Model Uncer...
 
多変量解析の一般化
多変量解析の一般化多変量解析の一般化
多変量解析の一般化
 
CVPR2016 reading - 特徴量学習とクロスモーダル転移について
CVPR2016 reading - 特徴量学習とクロスモーダル転移についてCVPR2016 reading - 特徴量学習とクロスモーダル転移について
CVPR2016 reading - 特徴量学習とクロスモーダル転移について
 
NIPS2015 reading - Learning visual biases from human imagination
NIPS2015 reading - Learning visual biases from human imaginationNIPS2015 reading - Learning visual biases from human imagination
NIPS2015 reading - Learning visual biases from human imagination
 
CVPR2015 reading "Understainding image virality" (in Japanese)
CVPR2015 reading "Understainding image virality" (in Japanese)CVPR2015 reading "Understainding image virality" (in Japanese)
CVPR2015 reading "Understainding image virality" (in Japanese)
 
Computational models of human visual attention driven by auditory cues
Computational models of human visual attention driven by auditory cuesComputational models of human visual attention driven by auditory cues
Computational models of human visual attention driven by auditory cues
 
NIPS2014 reading - Top rank optimization in linear time
NIPS2014 reading - Top rank optimization in linear timeNIPS2014 reading - Top rank optimization in linear time
NIPS2014 reading - Top rank optimization in linear time
 
CVPR2014 reading "Reconstructing storyline graphs for image recommendation fr...
CVPR2014 reading "Reconstructing storyline graphs for image recommendation fr...CVPR2014 reading "Reconstructing storyline graphs for image recommendation fr...
CVPR2014 reading "Reconstructing storyline graphs for image recommendation fr...
 
ICCV2013 reading: Learning to rank using privileged information
ICCV2013 reading: Learning to rank using privileged informationICCV2013 reading: Learning to rank using privileged information
ICCV2013 reading: Learning to rank using privileged information
 
ACMMM 2013 reading: Large-scale visual sentiment ontology and detectors using...
ACMMM 2013 reading: Large-scale visual sentiment ontology and detectors using...ACMMM 2013 reading: Large-scale visual sentiment ontology and detectors using...
ACMMM 2013 reading: Large-scale visual sentiment ontology and detectors using...
 
IJCAI13 Paper review: Large-scale spectral clustering on graphs
IJCAI13 Paper review: Large-scale spectral clustering on graphsIJCAI13 Paper review: Large-scale spectral clustering on graphs
IJCAI13 Paper review: Large-scale spectral clustering on graphs
 
関西CVPR勉強会 2012.10.28
関西CVPR勉強会 2012.10.28関西CVPR勉強会 2012.10.28
関西CVPR勉強会 2012.10.28
 
関西CVPR勉強会 2012.7.29
関西CVPR勉強会 2012.7.29関西CVPR勉強会 2012.7.29
関西CVPR勉強会 2012.7.29
 
ICWSM12 Brief Review
ICWSM12 Brief ReviewICWSM12 Brief Review
ICWSM12 Brief Review
 
関西CVPRML勉強会 2012.2.18 (一般物体認識 - データセット)
関西CVPRML勉強会 2012.2.18 (一般物体認識 - データセット)関西CVPRML勉強会 2012.2.18 (一般物体認識 - データセット)
関西CVPRML勉強会 2012.2.18 (一般物体認識 - データセット)
 
関西CVPRML勉強会(特定物体認識) 2012.1.14
関西CVPRML勉強会(特定物体認識) 2012.1.14関西CVPRML勉強会(特定物体認識) 2012.1.14
関西CVPRML勉強会(特定物体認識) 2012.1.14
 
人間の視覚的注意を予測するモデル - 動的ベイジアンネットワークに基づく 最新のアプローチ -
人間の視覚的注意を予測するモデル - 動的ベイジアンネットワークに基づく 最新のアプローチ -人間の視覚的注意を予測するモデル - 動的ベイジアンネットワークに基づく 最新のアプローチ -
人間の視覚的注意を予測するモデル - 動的ベイジアンネットワークに基づく 最新のアプローチ -
 
IBIS2011 企画セッション「CV/PRで独自の進化を遂げる学習・最適化技術」 趣旨説明
IBIS2011 企画セッション「CV/PRで独自の進化を遂げる学習・最適化技術」 趣旨説明IBIS2011 企画セッション「CV/PRで独自の進化を遂げる学習・最適化技術」 趣旨説明
IBIS2011 企画セッション「CV/PRで独自の進化を遂げる学習・最適化技術」 趣旨説明
 
立命館大学 AMLコロキウム 2011.10.20
立命館大学 AMLコロキウム 2011.10.20立命館大学 AMLコロキウム 2011.10.20
立命館大学 AMLコロキウム 2011.10.20
 

Recently uploaded

How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonetsnaman860154
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsEnterprise Knowledge
 
Azure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAzure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAndikSusilo4
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure servicePooja Nehwal
 
Next-generation AAM aircraft unveiled by Supernal, S-A2
Next-generation AAM aircraft unveiled by Supernal, S-A2Next-generation AAM aircraft unveiled by Supernal, S-A2
Next-generation AAM aircraft unveiled by Supernal, S-A2Hyundai Motor Group
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Scott Keck-Warren
 
Key Features Of Token Development (1).pptx
Key  Features Of Token  Development (1).pptxKey  Features Of Token  Development (1).pptx
Key Features Of Token Development (1).pptxLBM Solutions
 
Artificial intelligence in the post-deep learning era
Artificial intelligence in the post-deep learning eraArtificial intelligence in the post-deep learning era
Artificial intelligence in the post-deep learning eraDeakin University
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking MenDelhi Call girls
 
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphSIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphNeo4j
 
Unblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesUnblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesSinan KOZAK
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersThousandEyes
 
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | DelhiFULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhisoniya singh
 
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Alan Dix
 
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsSnow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsHyundai Motor Group
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machinePadma Pradeep
 
Pigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreternaman860154
 

Recently uploaded (20)

How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonets
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
 
Azure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAzure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & Application
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
 
Next-generation AAM aircraft unveiled by Supernal, S-A2
Next-generation AAM aircraft unveiled by Supernal, S-A2Next-generation AAM aircraft unveiled by Supernal, S-A2
Next-generation AAM aircraft unveiled by Supernal, S-A2
 
Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024
 
Key Features Of Token Development (1).pptx
Key  Features Of Token  Development (1).pptxKey  Features Of Token  Development (1).pptx
Key Features Of Token Development (1).pptx
 
Artificial intelligence in the post-deep learning era
Artificial intelligence in the post-deep learning eraArtificial intelligence in the post-deep learning era
Artificial intelligence in the post-deep learning era
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
 
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphSIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
 
Vulnerability_Management_GRC_by Sohang Sengupta.pptx
Vulnerability_Management_GRC_by Sohang Sengupta.pptxVulnerability_Management_GRC_by Sohang Sengupta.pptx
Vulnerability_Management_GRC_by Sohang Sengupta.pptx
 
Unblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesUnblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen Frames
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
 
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | DelhiFULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
 
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
 
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter RoadsSnow Chain-Integrated Tire for a Safe Drive on Winter Roads
Snow Chain-Integrated Tire for a Safe Drive on Winter Roads
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machine
 
Pigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping Elbows
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreter
 

CVPR2015 reading "Global refinement of random forest"

  • 3. 1-page summary • A method for refining a pre-trained random forest – Comparable to RF with much more nodes of decision trees – Better than RF with the same size of decision trees 3
  • 4. Random forest • An ensemble of decision trees trained by bootstrap sampling & random feature selection Decision tree 𝒙𝒙𝑖𝑖 �𝑦𝑦𝑖𝑖 = 𝑓𝑓𝑡𝑡(𝒙𝒙𝑖𝑖) (returns a MAP prediction) 4
  • 5. Random forest • An ensemble of decision trees trained by bootstrap sampling & random feature selection 𝒙𝒙𝑖𝑖 �𝑦𝑦𝑖𝑖 = 𝑓𝑓1(𝒙𝒙𝑖𝑖) (returns a MAP prediction) 𝒙𝒙𝑖𝑖 �𝑦𝑦𝑖𝑖 = 𝑓𝑓2(𝒙𝒙𝑖𝑖) 𝒙𝒙𝑖𝑖 �𝑦𝑦𝑖𝑖 = 𝑓𝑓𝑇𝑇(𝒙𝒙𝑖𝑖) Random forest 5
  • 6. Reformulation of a decision tree • A prediction 𝑓𝑓 can be divided into 2 components Decision tree 𝒙𝒙𝑖𝑖 �𝑦𝑦𝑖𝑖 = 𝑓𝑓𝑡𝑡(𝒙𝒙𝑖𝑖) (returns a MAP prediction) 𝜙𝜙𝑡𝑡 𝒙𝒙𝑖𝑖 = 0, 0, 1, 0 𝑇𝑇 Indicating which path to a leaf node is selected, can be represented by a binary vector. 𝑤𝑤𝑡𝑡(𝑦𝑦) = 0.2, 0.5, 0.8, 0.1 𝑇𝑇 Storing a posterior probability of 𝑦𝑦 at each leaf node, can be represented by a real vector. 𝑓𝑓𝑡𝑡 𝒙𝒙𝑖𝑖 = arg max 𝑦𝑦 𝑤𝑤𝑡𝑡(𝑦𝑦) ⋅ 𝜙𝜙𝑡𝑡(𝒙𝒙𝑖𝑖) Indicator vector Leaf vector 6
  • 7. Reformulation of a random forest • A prediction 𝑓𝑓 can be divided into 2 components 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖 0.2 0.5 0.8 0.1 0.3 0.7 0.1 0.2 0.1 0.1 0.5 0.3𝑊𝑊 𝑦𝑦 Random forest 𝑓𝑓 𝒙𝒙𝑖𝑖 = arg max 𝑦𝑦 𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙𝑖𝑖) 7 Look like a SVM classifier
  • 8. Global refinement • Optimize a leaf vector (weights) 𝑊𝑊(𝑦𝑦), while maintaining the indicator vector (structure) Φ(𝑥𝑥) 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖 0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦 Random forest 𝑓𝑓 𝒙𝒙𝑖𝑖 = arg max 𝑦𝑦 𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙𝑖𝑖) 8
  • 9. Global refinement • Optimize a leaf vector (weights) 𝑊𝑊(𝑦𝑦), while maintaining the indicator vector (structure) Φ(𝑥𝑥) 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖 0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦 Random forest 𝑓𝑓 𝒙𝒙𝑖𝑖 = arg max 𝑦𝑦 𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙𝑖𝑖) This optimization can be regarded as a linear classification problem, where an indicator vector Φ(𝒙𝒙) is a new representation of a sample 𝒙𝒙. [Note] In standard random forest, the trees are independently optimized. This optimization effectively utilizes complementary information among trees. 9
  • 10. Global refinement • Optimize a leaf vector (weights) 𝑊𝑊(𝑦𝑦), while maintaining the indicator vector (structure) Φ(𝑥𝑥) 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖 0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦 Random forest 𝑓𝑓 𝒙𝒙𝑖𝑖 = arg max 𝑦𝑦 𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙𝑖𝑖) This optimization can be regarded as a linear classification problem, where an indicator vector Φ(𝒙𝒙) is a new representation of a sample 𝒙𝒙. A sample Φ(𝑥𝑥) is highly sparse  Liblinear well suits this problem. It can be easily extended to a regression problem. 10
  • 11. Global pruning • Adjacent leaves with nearly-zero weights 𝑊𝑊(𝑦𝑦) do not contribute to the final result.  merge them. 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖 0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦 Random forest 𝑓𝑓 𝒙𝒙 = arg max 𝑦𝑦 𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙) 0 0.1 0 0.1 11
  • 12. Global pruning • Adjacent leaves with nearly-zero weights 𝑊𝑊(𝑦𝑦) do not contribute to the final result.  merge them. 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 𝒙𝒙𝑖𝑖 0 0 1 0 0 1 0 0 0 0 1 0Φ 𝒙𝒙𝑖𝑖 0.1 0.3 0.9 0.1 0.1 0.8 0.1 0.2 0.1 0.1 0.7 0.1�𝑊𝑊 𝑦𝑦 Random forest 𝑓𝑓 𝒙𝒙 = arg max 𝑦𝑦 𝑊𝑊(𝑦𝑦) ⋅ Φ(𝒙𝒙) 0 0.1 0 0.1 1. Optimize leaf vectors 𝑊𝑊 𝑦𝑦 ∀𝑦𝑦 2. Prune a certain percentage of insignificant leaves (significance = sum of elements in leaf vectors) 3. Update indicator vectors Φ(𝑥𝑥) for all the training samples 4. Repeat 1-3 until satisfying certain criterion, e.g. a. the size of the random forest is smaller than predefined, b. the prediction accuracy achieves best on a validation set. 12
  • 13. Data sets for experiments 13
  • 14. Experimental results • ADF/ARF - alternating decision (regression) forest [Schulter+ ICCV13] • Refined-A - Proposed method with the “accuracy” criterion • Refined-E - Proposed method with “over-pruning” (Accuracy is comparable to the original RF, but the size is much smaller.) • Metrics - Error rate for classification, RMSE for regression. • # trees = 100, max. depth = 10, 15 or 25 depending on the size of the training data. • 60% for training, 40% for testing. 14
  • 15. Parameter analysis • The proposed method achieved better performances than RFs with the same tree parameters (e.g. the number and depth of trees) 15 (for MNIST data)
  • 16. Parameter analysis • The proposed method accelerates both training and testing steps 16 (for MNIST data) Number of dimensions used on each node splitting Number of samples used in each decision tree Best for RFBest for proposed Time for testingfast slow Time for trainingfast slow Less sensitive More samples needed
  • 17. Applications • Kinect body part classification • Age regression from face images 17 Task-specific features 
  • 18. Last words • Simple, easy to implement, but effective • Can be applicable to other classifiers 18