cost sensitive dynamic feature selection
play

Cost-sensitive Dynamic Feature Selection e III 1 and Jason Eisner 2 - PowerPoint PPT Presentation

Nov 05, 2023 •176 likes •641 views

Cost-sensitive Dynamic Feature Selection e III 1 and Jason Eisner 2 He He 1 , Hal Daum 1 University of Maryland, College Park 2 Johns Hopkins University June 30, 2012 He He, Hal Daum e III and Jason Eisner Cost-sensitive Dynamic Feature

  1. Cost-sensitive Dynamic Feature Selection e III 1 and Jason Eisner 2 He He 1 , Hal Daum´ 1 University of Maryland, College Park 2 Johns Hopkins University June 30, 2012 He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 1 / 14

  2. Dynamic Feature Selection Feature selection in real life is a sequential decision-making process. He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 2 / 14

  3. Dynamic Feature Selection Feature selection in real life is a sequential decision-making process. Observation cost Decision coughing free cold flu H1N1 cold flu sore throat free H1N1 flu H1N1 headache free cold He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 2 / 14

  4. Dynamic Feature Selection Feature selection in real life is a sequential decision-making process. Observation cost Decision coughing free cold flu H1N1 cold flu sore throat free H1N1 flu H1N1 headache free cold flu H1N1 temperature (101 ◦ ) $1 cold H1N1 nasal swab test (pos.) $10 flu cold H1N1 viral culture test (pos.) $50 cold flu H1N1 molecular test (pos.) $100 cold flu H1N1 blood test (pos.) $100 cold flu He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 2 / 14

  5. Dynamic Feature Selection Feature selection in real life is a sequential decision-making process. Observation cost Decision coughing free cold flu H1N1 cold flu sore throat free H1N1 flu H1N1 headache free cold flu H1N1 temperature (101 ◦ ) $1 cold H1N1 nasal swab test (pos.) $10 flu cold H1N1 viral culture test (pos.) $50 cold flu H1N1 molecular test (pos.) $100 cold flu H1N1 blood test (pos.) $100 cold flu He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 2 / 14

  6. Dynamic Feature Selection Feature selection in real life is a sequential decision-making process. Observation cost Decision coughing free cold flu H1N1 cold flu sore throat free H1N1 flu H1N1 headache free cold flu H1N1 temperature (101 ◦ ) $1 cold flu nasal swab test (neg.) $10 H1N1 cold He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 2 / 14

  7. Dynamic Feature Selection Feature selection in real life is a sequential decision-making process. Observation cost Decision coughing free cold flu H1N1 cold flu sore throat free H1N1 flu H1N1 headache free cold flu H1N1 temperature (101 ◦ ) $1 cold flu nasal swab test (neg.) $10 H1N1 cold H1N1 viral culture test (pos.) $50 flu cold H1N1 molecular test (pos.) $100 cold flu H1N1 blood test (pos.) $100 cold flu He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 2 / 14

  8. Cost-sensitive Dynamic Feature Selection Feature Cost • Computation time • Data acquisition expense Dynamic Selection • Based on previous selected features and their values • Compute features on-the-run Given a pretrained classifier and feature cost, Goal • Sequentially select features for each instance at test time • Achieve a user-specified accuracy-cost trade-off He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 3 / 14

  9. Dynamic Feature Selection as an MDP He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 4 / 14

  10. Dynamic Feature Selection as an MDP At time step t , for one example, State s t Selected features and their values Action a t ∈ A t Acquire some features or stop Policy π Map from state to action: π ( s t ) = a t Reward r r ( s t , a t ) = margin( s t , a t ) − λ · cost( s t , a t ) margin : score of the true class - highest score of other classes λ : trade-off parameter He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 4 / 14

  11. Imitation Learning Oracle • Demonstrate optimal actions π ∗ ( s ) = a ∗ t Agent • Learn a policy to mimic the oracle’s behavior • π ( s t ) = a t Imitation via Supervised Classification • Training examples { ( φ ( s π ∗ ) , π ∗ ( s )) } • Feature: φ ( s ) label: π ∗ ( s ) classifier: ˆ π • Minimize a surrogate loss ℓ ( s , π ) w.r.t. to π ∗ , e.g. hinge loss in SVM. He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 5 / 14

  12. Forward Selection Oracle • Select the feature that yields the maximum immediate reward He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  13. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  14. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 temperature nasal swab test viral culture test molecular test blood test He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  15. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 temperature (101 ◦ ) -0.10 0.01 -0.11 nasal swab test viral culture test molecular test blood test He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  16. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 temperature nasal swab test (pos.) 0.50 0.04 0.46 viral culture test molecular test blood test He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  17. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 temperature nasal swab test viral culture test (pos.) 0.60 0.19 0.41 molecular test blood test He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  18. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 temperature nasal swab test viral culture test molecular test (pos.) 0.70 0.38 0.32 blood test He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  19. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 temperature nasal swab test viral culture test molecular test blood test (pos.) 0.65 0.38 0.27 He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  20. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 temperature 2 nasal swab test (pos.) 0.50 0.04 0.46 viral culture test molecular test blood test He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

  21. Forward Selection Oracle • Select the feature that yields the maximum immediate reward r ( s t , a t ) = margin( s t , a t ) − cost( s t , a t ) λ = 1, cost scaled to [0 , 1], H1N1=positive order feat. marg. cost reward 1 coughing, sore throat, headache -0.20 0.00 -0.10 3 temperature (101 ◦ ) 0.55 0.05 0.50 2 nasal swab test (pos.) 0.50 0.04 0.46 viral culture test molecular test blood test He He, Hal Daum´ e III and Jason Eisner Cost-sensitive Dynamic Feature Selection June 30, 2012 6 / 14

Recommend

Feature Selection: ROC and Subset Selection Theodoridis 5.5-5.7 Using ROC for Feature Selection

Feature Selection: ROC and Subset Selection Theodoridis 5.5-5.7 Using ROC for Feature Selection

Feature Selection: ROC and Subset Selection Theodoridis 5.5-5.7 Using ROC for Feature Selection Hypothesis Tests Examined (e.g. t-test): Useful for discarding features But does not tell us about overlap between classes for a feature! At Left

526 views • 21 slides
Dynamic Feature Selection for Dependency Parsing He He, Hal Daum III and Jason Eisner EMNLP

Dynamic Feature Selection for Dependency Parsing He He, Hal Daum III and Jason Eisner EMNLP

Dynamic Feature Selection for Dependency Parsing He He, Hal Daum III and Jason Eisner EMNLP 2013, Seattle Structured Prediction in NLP Part-of-Speech Tagging Parsing N N V Det N

1.59k views • 102 slides
Boosting Flexible Learning Ensembles with Dynamic Feature Selection Alexander Borisov, Victor

Boosting Flexible Learning Ensembles with Dynamic Feature Selection Alexander Borisov, Victor

Boosting Flexible Learning Ensembles with Dynamic Feature Selection Alexander Borisov, Victor Eruhimov, Eugene Tuv Intel Corp. Challenging models / data we face both regression and classification models are of interest mixed type

843 views • 14 slides
Outline Reducing Dimensionality Feature Selection 1 Steven J Zeil Feature Extraction 2

Outline Reducing Dimensionality Feature Selection 1 Steven J Zeil Feature Extraction 2

Feature Selection Feature Extraction Feature Selection Feature Extraction Outline Reducing Dimensionality Feature Selection 1 Steven J Zeil Feature Extraction 2 Principal Components Analysis (PCA) Old Dominion Univ. Factor Analysis (FA)

602 views • 6 slides
Reducing Dimensionality Steven J Zeil Old Dominion Univ. Fall 2010 1 Feature Selection

Reducing Dimensionality Steven J Zeil Old Dominion Univ. Fall 2010 1 Feature Selection

Feature Selection Feature Extraction Reducing Dimensionality Steven J Zeil Old Dominion Univ. Fall 2010 1 Feature Selection Feature Extraction Outline Feature Selection 1 Feature Extraction 2 Principal Components Analysis (PCA) Factor

479 views • 36 slides
Feature Selection ZHI LI Fenys Lab October 3, 2019 What is Feature? X (Independent)

Feature Selection ZHI LI Fenys Lab October 3, 2019 What is Feature? X (Independent)

Feature Selection ZHI LI Fenys Lab October 3, 2019 What is Feature? X (Independent) Variable Predictor Covariate Face; leg; tail; Hair texture/style Impeachment Trade war Whatever you can think of.. What methods are out there?

826 views • 37 slides
Mutual Information an Adequate Tool for Feature Selection ? Benot Frnay November 15, 2013

Mutual Information an Adequate Tool for Feature Selection ? Benot Frnay November 15, 2013

Mutual Information an Adequate Tool for Feature Selection ? Benot Frnay November 15, 2013 Introduction What is Feature Selection ? Overview of the Presentation Example of Feature Selection: Diabetes Progression Goal : predict the diabetes

1.05k views • 64 slides
Evolutionary Computation for Feature Selection and Feature Construction Bing Xue School of

Evolutionary Computation for Feature Selection and Feature Construction Bing Xue School of

Evolutionary Computation for Feature Selection and Feature Construction Bing Xue School of Engineering and Computer Science Victoria University of Wellington Bing.Xue@ecs.vuw.ac.nz IEEE CIS Webinar Mon, Sep 25, 2017 2:00 PM - 3:00 PM NZDT

1.7k views • 67 slides
Feature Selection CS 760@UW-Madison Goals for the lecture you should understand the following

Feature Selection CS 760@UW-Madison Goals for the lecture you should understand the following

Feature Selection CS 760@UW-Madison Goals for the lecture you should understand the following concepts filtering-based feature selection information gain filtering Markov blanket filtering frequency pruning

686 views • 25 slides
Feature Selection Richard Pospesel and Bert Wierenga Introduction Preprocessing Peaking

Feature Selection Richard Pospesel and Bert Wierenga Introduction Preprocessing Peaking

Feature Selection Richard Pospesel and Bert Wierenga Introduction Preprocessing Peaking Phenomenon Feature Selection Based on Statistical Hypothesis T esting Dimensionality Reduction Using Neural Networks Outlier Removal

642 views • 17 slides
Feature Selection CE-725: Statistical Pattern Recognition Sharif University of Technology

Feature Selection CE-725: Statistical Pattern Recognition Sharif University of Technology

Feature Selection CE-725: Statistical Pattern Recognition Sharif University of Technology Soleymani Fall 2018 Some slides are based on Isabelle Guyons slides : Introduction to feature selection, 2007. Outline } Dimensionality

1.08k views • 34 slides
RIDGE and LASSO regularization for regression Feature selection - Some algorithms perform

RIDGE and LASSO regularization for regression Feature selection - Some algorithms perform

RIDGE and LASSO regularization for regression Feature selection - Some algorithms perform naturally feature selection - for example Decision Trees, Boosting - Other algorithms have difficulty with correlated features - for example Naive Bayes,

801 views • 14 slides
Week 3 Video 4 Automated Feature Generation Automated Feature Selection Automated Feature

Week 3 Video 4 Automated Feature Generation Automated Feature Selection Automated Feature

Week 3 Video 4 Automated Feature Generation Automated Feature Selection Automated Feature Generation The creation of new data features in an automated fashion from existing data features Multiplicative Interactions You have variables A

513 views • 33 slides
Causal and Non-Causal Feature Selection for Ridge Regression Gavin Cawley School of Computing

Causal and Non-Causal Feature Selection for Ridge Regression Gavin Cawley School of Computing

Causal and Non-Causal Feature Selection for Ridge Regression Gavin Cawley School of Computing Sciences University of East Anglia Norwich, United Kingdom gcc@cmp.uea.ac.uk Wednesday 3 rd June 2008 Introduction Causal feature selection

246 views • 13 slides
Dynamic Classifier Selection Based on Imprecise Probabilities Meizhu Li Ghent University

Dynamic Classifier Selection Based on Imprecise Probabilities Meizhu Li Ghent University

Dynamic Classifier Selection Based on Imprecise Probabilities Meizhu Li Ghent University Co-work: Jasper De Bock, Gert de Cooman Outline Dynamic Strategy of Experiment classifier selection results selection Motivation Normally, one

354 views • 18 slides
10601 Machine Learning Model and feature selection Model selection issues We have seen some

10601 Machine Learning Model and feature selection Model selection issues We have seen some

10601 Machine Learning Model and feature selection Model selection issues We have seen some of this before Selecting features (or basis functions) Logistic regression SVMs Selecting parameter value Prior strength

611 views • 31 slides
... Other types of activation functions ( net = w i x i ) x n w n 1 1 x i : i- th

... Other types of activation functions ( net = w i x i ) x n w n 1 1 x i : i- th

M otivation Knowledge discovery process FEATURE SELECTION USING Interpretation ANT COLONY OPTIM IZATION: M odeling Knowledge APPLICATIONS IN HEALTH CARE Feature selection Patterns Preprocessing Reduced data Joo M . C. Sousa 1 Data

159 views • 14 slides
Principal Component Analysis (PCA) CE-717: Machine Learning Sharif University of Technology

Principal Component Analysis (PCA) CE-717: Machine Learning Sharif University of Technology

Principal Component Analysis (PCA) CE-717: Machine Learning Sharif University of Technology Spring 2016 Soleymani Dimensionality Reduction: Feature Selection vs. Feature Extraction Feature selection Select a subset of a given feature

5.23k views • 40 slides
Mach Machine Le ine Learning arning Feature Space, Feature Selection Hamid R. Rabiee Jafar

Mach Machine Le ine Learning arning Feature Space, Feature Selection Hamid R. Rabiee Jafar

Mach Machine Le ine Learning arning Feature Space, Feature Selection Hamid R. Rabiee Jafar Muhammadi Spring 2015 http://ce.sharif.edu/courses/93-94/2/ce717-1 / Agenda Agenda Features and Patterns The Curse of Size and

731 views • 27 slides
7100 Plus Highlights Elegant and stylish design New touch sensitive selection panel

7100 Plus Highlights Elegant and stylish design New touch sensitive selection panel

7100 Plus Highlights Elegant and stylish design New touch sensitive selection panel Wide selection of drinks Patented Sigma brewer for perfect fresh brew tea and coffee based drinks Easy to maintain thanks to colour

488 views • 9 slides
PCA & ICA CE-717: Machine Learning Sharif University of Technology Spring 2018 Soleymani

PCA & ICA CE-717: Machine Learning Sharif University of Technology Spring 2018 Soleymani

PCA & ICA CE-717: Machine Learning Sharif University of Technology Spring 2018 Soleymani Dimensionality Reduction: Feature Selection vs. Feature Extraction } Feature selection } Select a subset of a given feature set } Feature extraction }

875 views • 51 slides
Data Warehousing and Machine Learning Feature Selection Thomas D. Nielsen Aalborg University

Data Warehousing and Machine Learning Feature Selection Thomas D. Nielsen Aalborg University

Data Warehousing and Machine Learning Feature Selection Thomas D. Nielsen Aalborg University Department of Computer Science Spring 2008 DWML Spring 2008 1 / 16 Feature Selection When features dont help Data generated by process

578 views • 31 slides
Boosting Convergence of Timing Closure using Feature Selection in a Learning-driven Approach Que

Boosting Convergence of Timing Closure using Feature Selection in a Learning-driven Approach Que

Boosting Convergence of Timing Closure using Feature Selection in a Learning-driven Approach Que Yanghua, Harnhua Ng, Nachiket Kapre yanghua.que@ntu.edu.sg, nachiket@ieee.org Claim 2 Claim Feature Selection helps boost AUC scores for

1.09k views • 72 slides
Feature Import Vector Machine (FIVM): A General Classifier with Flexible Feature Selection

Feature Import Vector Machine (FIVM): A General Classifier with Flexible Feature Selection

Feature Import Vector Machine (FIVM): A General Classifier with Flexible Feature Selection This is a joint work with Y. Wang

489 views • 36 slides

More recommend