a multitask multiple kernel learning algorithm for
play

A Multitask Multiple Kernel Learning Algorithm for Survival Analysis - PowerPoint PPT Presentation

Aug 15, 2023 •185 likes •341 views

A Multitask Multiple Kernel Learning Algorithm for Survival Analysis with Application to Cancer Biology Onur Dereli, Ceyda O guz, Mehmet Gnen Ko University, Istanbul, Turkey odereli14@ku.edu.tr coguz@ku.edu.tr mehmetgonen@ku.edu.tr

  1. A Multitask Multiple Kernel Learning Algorithm for Survival Analysis with Application to Cancer Biology Onur Dereli, Ceyda O˘ guz, Mehmet Gönen Koç University, ˙ Istanbul, Turkey odereli14@ku.edu.tr coguz@ku.edu.tr mehmetgonen@ku.edu.tr June 12, 2019 / Long Beach Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 1 / 9

  2. multiple kernel learning Vital follow-up Genes Patients Genes Patients death status Days to last Days to Patients Patients Patients 456 Patients analysis Survival analysis Patients Patients Survival Patients Patients Multitask Patients Patients Dead Patients Patients . NA 1891 Dead 2208 NA Alive . . . . . . Genes . . NA 3200 Dead NA Genes Patients Genes Patients Genes Patients Patients Proposed Approach Vital Days to Days to last X 1 , 1 K 1 , 1 status death follow-up η 1 Alive NA 678 f 1 · · · · · · · · · Dead 364 NA . . . Gene set 1 η P . . . . . . X 1 ,P K 1 ,P G1 Alive NA 2555 K 1 ,η Dead 520 NA X 1 G17 G42 Y 1 G28 G6 · · · G3 G19 · · · · · · · · · · · · · · · · · · G42 G19 · · · Y T G8 G47 X T K T,η X T, 1 K T, 1 G12 G25 η 1 Gene set P f T · · · · · · · · · η P X T,P K T,P Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 2 / 9

  3. multiple kernel learning Vital follow-up Genes Patients Genes Patients death status Days to last Days to Patients Patients Patients 456 Patients analysis Survival analysis Patients Patients Survival Patients Patients Multitask Patients Patients Dead Patients Patients . NA 1891 Dead 2208 NA Alive . . . . . . Genes . . NA 3200 Dead NA Genes Patients Genes Patients Genes Patients Patients Proposed Approach Vital Days to Days to last X 1 , 1 K 1 , 1 status death follow-up η 1 Alive NA 678 f 1 · · · · · · · · · Dead 364 NA . . . Gene set 1 η P . . . . . . X 1 ,P K 1 ,P G1 Alive NA 2555 X 1 K 1 ,η Dead 520 NA G17 G42 Y 1 G28 G6 · · · G3 G19 · · · · · · · · · · · · · · · · · · G42 G19 · · · Y T G8 G47 K T,η X T X T, 1 K T, 1 G12 G25 η 1 Gene set P f T · · · · · · · · · η P X T,P K T,P Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 2 / 9

  4. multiple kernel learning Vital follow-up Genes Patients Genes Patients death status Days to last Days to Patients Patients Patients 456 Patients analysis Survival analysis Patients Patients Survival Patients Patients Multitask Patients Patients Dead Patients Patients . NA 1891 Dead 2208 NA Alive . . . . . . Genes . . NA 3200 Dead NA Genes Patients Genes Patients Genes Patients Patients Proposed Approach Vital Days to Days to last X 1 , 1 K 1 , 1 status death follow-up η 1 Alive NA 678 f 1 · · · · · · · · · Dead 364 NA . . . Gene set 1 η P . . . . . . X 1 ,P K 1 ,P G1 Alive NA 2555 X 1 K 1 ,η Dead 520 NA G17 G42 Y 1 G28 G6 · · · G3 G19 · · · · · · · · · · · · · · · · · · G42 G19 · · · Y T G8 G47 K T,η X T X T, 1 K T, 1 G12 G25 η 1 Gene set P f T · · · · · · · · · η P X T,P K T,P Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 2 / 9

  5. multiple kernel learning Vital follow-up Genes Patients Genes Patients death status Days to last Days to Patients Patients Patients 456 Patients analysis Survival analysis Patients Patients Survival Patients Patients Multitask Patients Patients Dead Patients Patients . NA 1891 Dead 2208 NA Alive . . . . . . Genes . . NA 3200 Dead NA Genes Patients Genes Patients Genes Patients Patients Proposed Approach Vital Days to Days to last X 1 , 1 K 1 , 1 status death follow-up η 1 Alive NA 678 f 1 · · · · · · · · · Dead 364 NA . . . Gene set 1 η P . . . . . . X 1 ,P K 1 ,P G1 Alive NA 2555 K 1 ,η Dead 520 NA X 1 G17 G42 Y 1 G28 G6 · · · G3 G19 · · · · · · · · · · · · · · · · · · G42 G19 · · · Y T G8 G47 X T K T,η X T, 1 K T, 1 G12 G25 η 1 Gene set P f T · · · · · · · · · η P X T,P K T,P Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 2 / 9

  6. Multitask Survival MKL Formulation N t T � � 1 � � ( ξ + minimize 2 w ⊤ ti + ( 1 − δ ti ) ξ − t w t + C ti ) t = 1 i = 1 with respect to w t ∈ R D t , ξ + t ∈ R N t , ξ − t ∈ R N t , b t ∈ R subject to ǫ + ξ + ti ≥ y ti − w ⊤ t x ti − b t ∀ ( t , i ) ǫ + ξ − ti ≥ w ⊤ t x ti + b t − y ti ∀ ( t , i ) ξ + ti ≥ 0 ∀ ( t , i ) ξ − ti ≥ 0 ∀ ( t , i ) Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 3 / 9

  7. Multitask Survival MKL Formulation N t N t T T � � � � minimize − y ti ( α + ( α + ti − α − ti ) + ǫ ti + α − ti ) t = 1 i = 1 t = 1 i = 1 T N t N t P + 1 � � � � ( α + ti )( α + tj ) η m k m ( x ti , x tj ) ti − α − tj − α − 2 t = 1 i = 1 j = 1 m = 1 with respect to α + t ∈ R N t , α − t ∈ R N t , η ∈ R P N t � ( α + subject to ti ) = 0 ∀ t ti − α − i = 1 C ≥ α + ti ≥ 0 ∀ ( t , i ) C ( 1 − δ ti ) ≥ α − ti ≥ 0 ∀ ( t , i ) P � η m = 1 m = 1 η m ≥ 0 ∀ m Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 3 / 9

  8. Multitask Survival MKL Formulation � T N t N t η ( s ) α ( s ) ti α ( s ) � � � tj k m ( x ti , x tj ) m t = 1 i = 1 j = 1 η ( s + 1 ) = ∀ m m � T P N t N t η ( s ) α ( s ) ti α ( s ) � � � � tj k o ( x ti , x tj ) o t = 1 o = 1 i = 1 j = 1 Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 3 / 9

  9. Data Sets BLCA BRCA CESC COAD ESCA GBM HNSC KIRC KIRP LAML LGG LIHC LUAD LUSC OV PAAD READ SARC STAD UCEC ● (402) (1067) (291) (433) (160) (152) (498) (526) (285) (130) (506) (365) (500) (493) (372) (176) (156) (256) (348) (539) • 20 cancer data sets from TCGA database Gene expression profiles and survival characteristics • Hallmark Gene Set [1] & PID Pathway [2] Collections Dereli, O., O˘ guz, C., Gönen, M. (KU) ICML 2019 June 12, 2019 / Long Beach 4 / 9

Recommend

Multitask Learning Lei Tang Arizona State University Nov. 6th, 2006 Lei Tang Multitask

Multitask Learning Lei Tang Arizona State University Nov. 6th, 2006 Lei Tang Multitask

Multitask Learning Lei Tang Arizona State University Nov. 6th, 2006 Lei Tang Multitask Learning 1 Introduction 2 Typical applications 3 What is multitask Learning 4 Why multitask learning makes sense 5 Multitask Learning methods 6 Pros and Cons

419 views • 41 slides
Efficient Multiple Kernel Learning Lei Tang Outline What is Kernel Learning? Whats the

Efficient Multiple Kernel Learning Lei Tang Outline What is Kernel Learning? Whats the

Efficient Multiple Kernel Learning Lei Tang Outline What is Kernel Learning? Whats the problem with existing formulation? Two new formulations for large scale kernel selection selection SIL formulation (Cutting Planes)

345 views • 16 slides
Multiple Kernel Learning and Feature Space Denoising Fei Yan, Josef Kittler and Krystian

Multiple Kernel Learning and Feature Space Denoising Fei Yan, Josef Kittler and Krystian

Overview Kernel Methods Multiple Kernel Learning MKL and Feature Space Denoising Conclusions Multiple Kernel Learning and Feature Space Denoising Fei Yan, Josef Kittler and Krystian Mikolajczyk eNTERFACE10 Multiple Kernel Learning and

431 views • 39 slides
Consistent Multitask Learning with Nonlinear Output Constraints Carlo Ciliberto Department of

Consistent Multitask Learning with Nonlinear Output Constraints Carlo Ciliberto Department of

Consistent Multitask Learning with Nonlinear Output Constraints Carlo Ciliberto Department of Computer Science, UCL joint work w/ Alessandro Rudi, Lorenzo Rosasco and Massi Pontil Multitask Learning (MTL) MTL Mantra : leverage the similarities

329 views • 22 slides
A majorization-minimization algorithm for (multiple) hyperparameter learning Chuan-Sheng Foo

A majorization-minimization algorithm for (multiple) hyperparameter learning Chuan-Sheng Foo

A majorization-minimization algorithm for (multiple) hyperparameter learning Chuan-Sheng Foo Chuong B. Do Andrew Y. Ng Stanford University ICML 2009 Montreal, Canada 17 th June 2009 Supervised learning Training set of m IID examples

221 views • 21 slides
1 Boosting: Basic Algorithm AdaBoost Pseudocode TrainAdaBoost(D, BaseLearn) General Loop:

1 Boosting: Basic Algorithm AdaBoost Pseudocode TrainAdaBoost(D, BaseLearn) General Loop:

Learning Ensembles Learn multiple alternative definitions of a concept using different training data or different learning algorithms. Combine decisions of multiple definitions, e.g. using CS 391L: Machine Learning: weighted voting.

342 views • 3 slides
Learning From Data Lecture 25 The Kernel Trick Learning with only inner products The Kernel M.

Learning From Data Lecture 25 The Kernel Trick Learning with only inner products The Kernel M.

Learning From Data Lecture 25 The Kernel Trick Learning with only inner products The Kernel M. Magdon-Ismail CSCI 4100/6100 recap: Large Margin is Better Controling Overfitting Non-Separable Data 0.08 random hyperplane 2 w t w + C N E

576 views • 18 slides
An ensemble-based kernel le learning framework to handle data assimilation problems wit ith im

An ensemble-based kernel le learning framework to handle data assimilation problems wit ith im

An ensemble-based kernel le learning framework to handle data assimilation problems wit ith im imperfect forw rward simulators Xiaodong Luo, NORCE Outline Background and motivation An ensemble-based kernel algorithm for supervised

516 views • 35 slides
Kernel Learning with a Million Kernels Ashesh Jain SVN Vishwanathan IIT Delhi Purdue

Kernel Learning with a Million Kernels Ashesh Jain SVN Vishwanathan IIT Delhi Purdue

Kernel Learning with a Million Kernels Ashesh Jain SVN Vishwanathan IIT Delhi Purdue University Manik Varma Microsoft Research India Kernel Learning The objective in kernel learning is to jointly learn both SVM and kernel parameters

1.34k views • 38 slides
Not to Cry Wolf: Distantly Supervised Multitask Learning in Critical Care Patrick Schwab 1

Not to Cry Wolf: Distantly Supervised Multitask Learning in Critical Care Patrick Schwab 1

Not to Cry Wolf: Distantly Supervised Multitask Learning in Critical Care Patrick Schwab 1 @schwabpa Emanuela Keller 2 , Carl Muroi 2 , David J. Mack 2 , Christian Strssle 2 and Walter Karlen 1 1 Institute of Robotics and Intelligent Systems,

527 views • 34 slides
Speaker and Emotion Recognition of TV-Series Data Using Multimodal and Multitask Deep Learning

Speaker and Emotion Recognition of TV-Series Data Using Multimodal and Multitask Deep Learning

Speaker and Emotion Recognition of TV-Series Data Using Multimodal and Multitask Deep Learning Sashi Novitasari 1 , Quoc Truong Do 1 , Sakriani Saktj 1,3 , Dessi Lestari 2 , Satoshi Nakamura 1,3 1 Graduate School of Informatjon Science, Nara

395 views • 16 slides
Multilingual and Multitask Learning in seq2seq Models CMSC 470 Marine Carpuat Multilingual

Multilingual and Multitask Learning in seq2seq Models CMSC 470 Marine Carpuat Multilingual

Multilingual and Multitask Learning in seq2seq Models CMSC 470 Marine Carpuat Multilingual Machine Translation Neural MT only helps in high-resource settings Ongoing research Learn from other sources of supervision than pairs (E,F)

622 views • 32 slides
Multitask radiological modality invariant landmark localization using deep reinforcement learning

Multitask radiological modality invariant landmark localization using deep reinforcement learning

Multitask radiological modality invariant landmark localization using deep reinforcement learning Vishwa S. Parekh, Alex E. Bocchieri, Vladimir Braverman, Michael A. Jacobs The Russell H. Morgan Department of Radiology and Radiological Science,

464 views • 17 slides
A Multitask Learning Approach to Assess the Dysarthria Severity in Parkinsons Patients Juan

A Multitask Learning Approach to Assess the Dysarthria Severity in Parkinsons Patients Juan

A Multitask Learning Approach to Assess the Dysarthria Severity in Parkinsons Patients Juan Camilo Vsquez-Correa 1 , 2 , Toms Arias-Vergara 1 , 2 , 3 Juan Rafael Orozco-Arroyave 1 , 2 , and Elmar Nth 1 , 2 1 Faculty of Engineering,

487 views • 32 slides
Joint Multitask Learning for Community Question Answering Using Task-Specific Embeddings Shafiq

Joint Multitask Learning for Community Question Answering Using Task-Specific Embeddings Shafiq

Joint Multitask Learning for Community Question Answering Using Task-Specific Embeddings Shafiq Joty Llus Mrquez Preslav Nakov Nanyang Technological Core AI-Barcelona Qatar Computing University Amazon Research Institute, HBKU

472 views • 24 slides
Active Multitask Learning Using Both Supervised and Latent Shared Topics Ayan Acharya , Raymond J.

Active Multitask Learning Using Both Supervised and Latent Shared Topics Ayan Acharya , Raymond J.

Background Act-DSLDA & Act-NPDSLDA Datasets & Empirical Results References Active Multitask Learning Using Both Supervised and Latent Shared Topics Ayan Acharya , Raymond J. Mooney, Joydeep Ghosh UT Austin, Dept. of ECE & CS April

600 views • 32 slides
Distant-supervised Heterogeneous multitask learning for social event forecasting with

Distant-supervised Heterogeneous multitask learning for social event forecasting with

Distant-supervised Heterogeneous multitask learning for social event forecasting with multilingual indicators Liang Zhao George Mason University What are Spatiotemporal Events? Week 47 Week 46 Week 45 Epidemics outbreak on Week 47 ending

152 views • 11 slides
. . : =L distance functions Possible Euclidean . Manhattan - other .

. . : =L distance functions Possible Euclidean . Manhattan - other .

Midterm the Since Topics based Instance and classification ] learning NN . regression Kernel . mgohssien ... ] ' , margin than Peru " learning , . .dyu, online approaches trick kernel kernel based ^ ,

345 views • 11 slides
Multitask Learning with Low-Level Auxiliary Tasks 1 Traditional automatic speech recognition

Multitask Learning with Low-Level Auxiliary Tasks 1 Traditional automatic speech recognition

for Speech Recognition Shubham Toshniwal , Hao Tang, Liang Lu, Karen Livescu Toyota Technological Institute at Chicago Multitask Learning with Low-Level Auxiliary Tasks 1 Traditional automatic speech recognition (ASR) systems are modular.

751 views • 32 slides
Extending Unsupervised Neural Image Compression With Supervised Multitask Learning David Tellez,

Extending Unsupervised Neural Image Compression With Supervised Multitask Learning David Tellez,

Extending Unsupervised Neural Image Compression With Supervised Multitask Learning David Tellez, Diederik Hoppener, Cornelis Verhoef, Dirk Grunhagen, Pieter Nierop, Michal Drozdzal, Jeroen van der Laak, Francesco Ciompi david.tellezm@gmail.com

302 views • 17 slides
Output Kernel Learning Methods Francesco Dinuzzo 1 Cheng Soon Ong 2 Kenji Fukumizu 3 1 Max Planck

Output Kernel Learning Methods Francesco Dinuzzo 1 Cheng Soon Ong 2 Kenji Fukumizu 3 1 Max Planck

Output Kernel Learning Methods Francesco Dinuzzo 1 Cheng Soon Ong 2 Kenji Fukumizu 3 1 Max Planck Institute for Intelligent Systems, T ubingen, Germany 2 NICTA, Melbourne 3 Institute of Statistical Mathematics, Japan Part I Learning multiple

706 views • 43 slides
menu del dia learning (7 slides) purushottam kar (iit kanpur) accelerated kernel learning

menu del dia learning (7 slides) purushottam kar (iit kanpur) accelerated kernel learning

accelerated kernel learning 1 purushottam kar department of computer science and engineering indian institute of technology kanpur november 27, 2012 1 joint work with harish c. karnick purushottam kar (iit kanpur) accelerated kernel learning

2.17k views • 175 slides
Supervised Learning Supervised learning algorithms require the presence of a teacher who

Supervised Learning Supervised learning algorithms require the presence of a teacher who

Backpropagation Learning Algorithm An algorithm for learning the weights in the network, given a training set of input-output pairs { , } The algorithm is based on gradient descent method. Architecture : Weight on

303 views • 9 slides
1 Kernel methods & optimization One example of a kernel that is frequently used in practice

1 Kernel methods & optimization One example of a kernel that is frequently used in practice

Machine Learning Class Notes 9-25-12 Prof. David Sontag 1 Kernel methods & optimization One example of a kernel that is frequently used in practice and which allows for highly non-linear discriminant functions is the Gaussian kernel,

241 views • 5 slides

More recommend