SlideShare une entreprise Scribd logo
1  sur  19
Télécharger pour lire hors ligne
SigOpt. Confidential.
Interactive Tradeoffs Between
Competing Offline Metrics with
Bayesian Optimization
KDD 2019
2nd Annual Workshop
Online and Offline Evaluation of Interactive Systems
Michael McCourt, Research Engineer, SigOpt
SigOpt. Confidential.
About me
● Research engineering at SigOpt
● Focus on applied Bayesian optimization
● PhD from Cornell
● Avid Cleveland Cavaliers fan
About SigOpt
● Leading software solution for parameter
optimization and model experimentation
● Customers in finance, trading, media,
technology, consulting, energy, industry
● Free version of our solution for academia
available at sigopt.com/edu
SigOpt. Confidential.
Abstract for KDD 2019
2nd Annual Workshop on Offline and Online Evaluation of Interactive Systems
Many real world applications (ML models, simulators, etc.) have multiple competing
metrics that define performance; these require practitioners to carefully consider
potential tradeoffs. However, assessing and ranking this tradeoff is nontrivial,
especially when the number of metrics is more than two. Often times, practitioners
scalarize the metrics into a single objective, e.g., using a weighted sum.
In this talk, we pose this problem as a constrained multi-objective optimization
problem. By setting and updating the constraints, we can efficiently explore only the
region of the Pareto efficient frontier of the model/system of most interest. We
motivate this problem with the application of an experimental design setting, where
we are trying to fabricate high performance glass substrate for solar cell panels.
3
SigOpt. Confidential.
Most Metrics are Impacted by Free Parameters
How can these free parameters be chosen?
Generally, these are chosen to yield good future performance.
• This discussion only covers offline metrics.
• Some of the elements apply in an online setting as well.
Given a computable metric defining future performance, a search can be conducted for the free parameters
yielding acceptable/optimal performance.
• In many circumstances, evaluating this performance metric is costly.
• Example: Train a classification model and evaluate a validation accuracy.
• Example: Use financial data from the past year for a trading strategy and evaluate its profit on last
month’s data.
4
SigOpt. Confidential.
Searching for Free Parameters Requires Efficiency
Intelligently searching a fixed domain
Many searches benefit from efficiently (actively) learning about the circumstances of the search.
• Active learning -- “Active learning is closely related to experimental design … is most often adaptive …
employs an oracle for data labelling … is usually used to learn a model for classification.” -- [Brochu et al
2010]
Two adjacent fields of research have evolved.
• Bayesian optimization -- “Bayesian optimization is a sequential model-based approach to [optimizing a
function].” -- [Shahriari et al, 2016]
• Active search -- “Active search is an active learning setting with the goal of identifying as many
members of a given class as possible under a labeling budget.” -- [Jiang et al, 2017]
How we conduct this active learning will greatly impact efficiency of the search.
5
SigOpt. Confidential.
Bayesian Optimization
A graphical depiction of the iterative process
6
Build a statistical model Build a statistical model
Choose a next point Choose a next point
SigOpt. Confidential.
Bayesian Optimization
Efficiently Optimize a Scalar Function
To quote [Frazier 2018]: Bayesian optimization (BayesOpt/BO) is a class of machine-learning-based
optimization methods focused on [maximizing/minimizing a function with] the following properties ...
• Typically the dimension d is less than 20.
• The objective function f is continuous, as is the domain (which is likely a d-dimensional rectangle).
• f is expensive to evaluate: e.g., time, money, access
• f is black-box: if lacks known special structure like concavity or linearity.
• When we evaluate f, we observe only f(x); that is, the optimization is gradient-free.
• f is often observed in the presence of noise.
• Our focus is on finding a global rather than local optimum.
7
SigOpt. Confidential.
Many Metrics may Contribute to Success
How do we execute under these ambiguous circumstances?
Defining/measuring future performance is imprecise.
• Generally, several metrics will contribute to a sense of future performance.
• Not all metrics are equally important -- some may only need to reach a threshold.
• The feasible performance and preferred interaction between metrics may not be known a priori.
Popular multiobjective optimization strategies are often population-based (and not often sample-efficient).
8
SigOpt. Confidential.
Multiobjective Bayesian Optimization
Balancing competing metrics to find the Pareto frontier
Adaptations to BO to search for the efficient frontier:
• Change the problem to an active search problem [Jiang et al, 2018].
• Search for diverse points near the efficient frontier.
• Scalarize the problem with linear combinations of the metrics [Knowles, 2006].
• Define a hypervolume based acquisition function [Hernandez-Lobato et al, 2016, Emmerich et al, 2016].
• Scalarization through prior beliefs [Astudillo, 2017].
Guiding points
• Users wanted to interactively update the search process.
• Users felt uncomfortable stating a priori preferences.
9
SigOpt. Confidential.
Multiobjective Bayesian Optimization
Our strategy
We apply a strategy similar to what was discussed in [Letham et al, 2019].
1. Model all metrics independently.
• Requires no prior beliefs on how metrics interact.
• Missing data removed on a per metric basis if unrecorded.
2. Expose the efficient frontier through constrained scalar optimization.
• Enforce user constraints when given.
• Iterate through sub constraints to better resolve efficient frontier, if desired.
• Consider different regions of the frontier when parallelism is possible.
3. Allow users to change constraints as the search progresses.
• Allow the problems/goals to evolve as the user’s understanding evolves.
10
Variation on
Expected
Improvement
SigOpt. Confidential.
One strategy can be to randomly apply constraints.
Multiobjective Bayesian Optimization
Our strategy
11
Note: There are GIFs that do not
show up in this version of the
presentation. For a copy that
includes them, please email
contact@sigopt.com
SigOpt. Confidential.
Another strategy can be to “walk” up and down the constraint domain.
Multiobjective Bayesian Optimization
Our strategy
12
Note: There are GIFs that do not
show up in this version of the
presentation. For a copy that
includes them, please email
contact@sigopt.com
SigOpt. Confidential.
It can help to alternate which metric the constraint is imposed on.
Multiobjective Bayesian Optimization
Our strategy
13
Note: There are GIFs that do not
show up in this version of the
presentation. For a copy that
includes them, please email
contact@sigopt.com
SigOpt. Confidential.
Users can enforce their own bounds to focus on the desired outcome.
Multiobjective Bayesian Optimization
Our strategy
14
Note: There are GIFs that do not
show up in this version of the
presentation. For a copy that
includes them, please email
contact@sigopt.com
SigOpt. Confidential.
Users can also update their own bounds as the experiment goes on.
Multiobjective Bayesian Optimization
Our strategy
15
Note: There are GIFs that do not
show up in this version of the
presentation. For a copy that
includes them, please email
contact@sigopt.com
SigOpt. Confidential.
Awesome Applications of Bayesian Optimization
Who is using, and can use, BO?
16
● ML/DL hyperparameter tuning [Snoek et al, 2012; Feurer et al, 2015; Kandasamy et al, 2018]
● Engineering system design [Mockus, 1989; Jones et al, 1998; Forrester et al, 2008]
● Drug design [Negoescu et al, 2011; Frazier and Wang, 2016]
● Material design [Packwood, 2017; Haghanifar et al, 2019]
● Model calibration [Shoemaker et al, 2007; Shi et al, 2013; Letham et al. 2019]
● Reinforcement learning [Lizotte, 2008; Brochu et al, 2010; Martinez-Cantin et al, 2018]
There are so many others!
SigOpt. Confidential.17
A Joint Collaboration with University of Pittsburgh
[Haghanifar et al, 2019]
Metrics
• Light transmission
• Clarity (low haze)
• Water resistance
Constraints updated on all
metrics during the search.
Note: There is a video that
does not show up in this
version of the presentation.
For a copy that includes this,
please email
contact@sigopt.com
SigOpt. Confidential.
Future Work
How can we improve this process?
When black-box constraints exist, how can we encourage our search to respect them?
• Hallucinate bad function values at points which violate the constraints.
• Attenuate the expected improvement by the probability of failure [Gelbart, 2015].
• Model the constraints and average out the noisy behavior [Letham et al, 2019].
• Model the Lagrangian [Picheny et al, 2016].
Question: Exactly how black-box/expensive are these constraints (or the objective)?
• We can adapt to expensive constraints but a cheap objective [Gramacy et al, 2106].
Question: Can we help focus on the important region using preferences?
• Joint work extending [Astudillo, 2017] with Raul and Peter.
18
SigOpt. Confidential.
Thank you for having me!

Contenu connexe

Similaire à Interactive Tradeoffs Between Competing Offline Metrics with Bayesian Optimization

Week 2 24 - 28 July 2023 (1).pptx
Week 2 24 - 28 July 2023 (1).pptxWeek 2 24 - 28 July 2023 (1).pptx
Week 2 24 - 28 July 2023 (1).pptx
XuanQin3
 

Similaire à Interactive Tradeoffs Between Competing Offline Metrics with Bayesian Optimization (20)

Analytics
AnalyticsAnalytics
Analytics
 
Lp assign
Lp assignLp assign
Lp assign
 
Experimentation at Scale
Experimentation at ScaleExperimentation at Scale
Experimentation at Scale
 
Better Living Through Analytics - Strategies for Data Decisions
Better Living Through Analytics - Strategies for Data DecisionsBetter Living Through Analytics - Strategies for Data Decisions
Better Living Through Analytics - Strategies for Data Decisions
 
Agile Technical Leadership
Agile Technical LeadershipAgile Technical Leadership
Agile Technical Leadership
 
Telemarketing prediction project
Telemarketing prediction projectTelemarketing prediction project
Telemarketing prediction project
 
The Total Economic ImpactTM (TEI) of Neo4j, Featuring Forrester
The Total Economic ImpactTM (TEI) of Neo4j, Featuring ForresterThe Total Economic ImpactTM (TEI) of Neo4j, Featuring Forrester
The Total Economic ImpactTM (TEI) of Neo4j, Featuring Forrester
 
Project feasibility study report
Project feasibility study report Project feasibility study report
Project feasibility study report
 
Presentation lecture 2nd quantitative techniques
Presentation lecture 2nd quantitative techniquesPresentation lecture 2nd quantitative techniques
Presentation lecture 2nd quantitative techniques
 
1530 track2 reid
1530 track2 reid1530 track2 reid
1530 track2 reid
 
Banglka.pptx
Banglka.pptxBanglka.pptx
Banglka.pptx
 
Module 5 - Data Science Methodology.pdf
Module 5 - Data Science Methodology.pdfModule 5 - Data Science Methodology.pdf
Module 5 - Data Science Methodology.pdf
 
Supply Chain Strategy Assessment
Supply Chain Strategy AssessmentSupply Chain Strategy Assessment
Supply Chain Strategy Assessment
 
Data driven; People based
Data driven; People basedData driven; People based
Data driven; People based
 
Pareto-Optimal Search-Based Software Engineering (POSBSE): A Literature Survey
Pareto-Optimal Search-Based Software Engineering (POSBSE): A Literature SurveyPareto-Optimal Search-Based Software Engineering (POSBSE): A Literature Survey
Pareto-Optimal Search-Based Software Engineering (POSBSE): A Literature Survey
 
Guide to competitive assessment
Guide to competitive assessment Guide to competitive assessment
Guide to competitive assessment
 
Using BAPO to apply structure follows strategy
Using BAPO to apply structure follows strategyUsing BAPO to apply structure follows strategy
Using BAPO to apply structure follows strategy
 
Content Solution Quick Start (June 2014)
Content Solution Quick Start (June 2014)Content Solution Quick Start (June 2014)
Content Solution Quick Start (June 2014)
 
Week 2 24 - 28 July 2023 (1).pptx
Week 2 24 - 28 July 2023 (1).pptxWeek 2 24 - 28 July 2023 (1).pptx
Week 2 24 - 28 July 2023 (1).pptx
 
Rapid Optimization Application Development Using Excel and Solver
Rapid Optimization Application Development Using Excel and SolverRapid Optimization Application Development Using Excel and Solver
Rapid Optimization Application Development Using Excel and Solver
 

Plus de SigOpt

Plus de SigOpt (20)

Optimizing BERT and Natural Language Models with SigOpt Experiment Management
Optimizing BERT and Natural Language Models with SigOpt Experiment ManagementOptimizing BERT and Natural Language Models with SigOpt Experiment Management
Optimizing BERT and Natural Language Models with SigOpt Experiment Management
 
Experiment Management for the Enterprise
Experiment Management for the EnterpriseExperiment Management for the Enterprise
Experiment Management for the Enterprise
 
Efficient NLP by Distilling BERT and Multimetric Optimization
Efficient NLP by Distilling BERT and Multimetric OptimizationEfficient NLP by Distilling BERT and Multimetric Optimization
Efficient NLP by Distilling BERT and Multimetric Optimization
 
Detecting COVID-19 Cases with Deep Learning
Detecting COVID-19 Cases with Deep LearningDetecting COVID-19 Cases with Deep Learning
Detecting COVID-19 Cases with Deep Learning
 
Metric Management: a SigOpt Applied Use Case
Metric Management: a SigOpt Applied Use CaseMetric Management: a SigOpt Applied Use Case
Metric Management: a SigOpt Applied Use Case
 
Tuning for Systematic Trading: Talk 3: Training, Tuning, and Metric Strategy
Tuning for Systematic Trading: Talk 3: Training, Tuning, and Metric StrategyTuning for Systematic Trading: Talk 3: Training, Tuning, and Metric Strategy
Tuning for Systematic Trading: Talk 3: Training, Tuning, and Metric Strategy
 
Tuning for Systematic Trading: Talk 2: Deep Learning
Tuning for Systematic Trading: Talk 2: Deep LearningTuning for Systematic Trading: Talk 2: Deep Learning
Tuning for Systematic Trading: Talk 2: Deep Learning
 
Tuning for Systematic Trading: Talk 1
Tuning for Systematic Trading: Talk 1Tuning for Systematic Trading: Talk 1
Tuning for Systematic Trading: Talk 1
 
Tuning Data Augmentation to Boost Model Performance
Tuning Data Augmentation to Boost Model PerformanceTuning Data Augmentation to Boost Model Performance
Tuning Data Augmentation to Boost Model Performance
 
Modeling at Scale: SigOpt at TWIMLcon 2019
Modeling at Scale: SigOpt at TWIMLcon 2019Modeling at Scale: SigOpt at TWIMLcon 2019
Modeling at Scale: SigOpt at TWIMLcon 2019
 
Tuning 2.0: Advanced Optimization Techniques Webinar
Tuning 2.0: Advanced Optimization Techniques WebinarTuning 2.0: Advanced Optimization Techniques Webinar
Tuning 2.0: Advanced Optimization Techniques Webinar
 
SigOpt at Ai4 Finance—Modeling at Scale
SigOpt at Ai4 Finance—Modeling at Scale SigOpt at Ai4 Finance—Modeling at Scale
SigOpt at Ai4 Finance—Modeling at Scale
 
Machine Learning Infrastructure
Machine Learning InfrastructureMachine Learning Infrastructure
Machine Learning Infrastructure
 
SigOpt at Uber Science Symposium - Exploring the spectrum of black-box optimi...
SigOpt at Uber Science Symposium - Exploring the spectrum of black-box optimi...SigOpt at Uber Science Symposium - Exploring the spectrum of black-box optimi...
SigOpt at Uber Science Symposium - Exploring the spectrum of black-box optimi...
 
SigOpt at O'Reilly - Best Practices for Scaling Modeling Platforms
SigOpt at O'Reilly - Best Practices for Scaling Modeling PlatformsSigOpt at O'Reilly - Best Practices for Scaling Modeling Platforms
SigOpt at O'Reilly - Best Practices for Scaling Modeling Platforms
 
SigOpt at GTC - Tuning the Untunable
SigOpt at GTC - Tuning the UntunableSigOpt at GTC - Tuning the Untunable
SigOpt at GTC - Tuning the Untunable
 
SigOpt at GTC - Reducing operational barriers to optimization
SigOpt at GTC - Reducing operational barriers to optimizationSigOpt at GTC - Reducing operational barriers to optimization
SigOpt at GTC - Reducing operational barriers to optimization
 
Lessons for an enterprise approach to modeling at scale
Lessons for an enterprise approach to modeling at scaleLessons for an enterprise approach to modeling at scale
Lessons for an enterprise approach to modeling at scale
 
Modeling at scale in systematic trading
Modeling at scale in systematic tradingModeling at scale in systematic trading
Modeling at scale in systematic trading
 
SigOpt at MLconf - Reducing Operational Barriers to Model Training
SigOpt at MLconf - Reducing Operational Barriers to Model TrainingSigOpt at MLconf - Reducing Operational Barriers to Model Training
SigOpt at MLconf - Reducing Operational Barriers to Model Training
 

Dernier

%+27788225528 love spells in Boston Psychic Readings, Attraction spells,Bring...
%+27788225528 love spells in Boston Psychic Readings, Attraction spells,Bring...%+27788225528 love spells in Boston Psychic Readings, Attraction spells,Bring...
%+27788225528 love spells in Boston Psychic Readings, Attraction spells,Bring...
masabamasaba
 
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
VictoriaMetrics
 
%+27788225528 love spells in Knoxville Psychic Readings, Attraction spells,Br...
%+27788225528 love spells in Knoxville Psychic Readings, Attraction spells,Br...%+27788225528 love spells in Knoxville Psychic Readings, Attraction spells,Br...
%+27788225528 love spells in Knoxville Psychic Readings, Attraction spells,Br...
masabamasaba
 
Abortion Pill Prices Boksburg [(+27832195400*)] 🏥 Women's Abortion Clinic in ...
Abortion Pill Prices Boksburg [(+27832195400*)] 🏥 Women's Abortion Clinic in ...Abortion Pill Prices Boksburg [(+27832195400*)] 🏥 Women's Abortion Clinic in ...
Abortion Pill Prices Boksburg [(+27832195400*)] 🏥 Women's Abortion Clinic in ...
Medical / Health Care (+971588192166) Mifepristone and Misoprostol tablets 200mg
 
%+27788225528 love spells in Toronto Psychic Readings, Attraction spells,Brin...
%+27788225528 love spells in Toronto Psychic Readings, Attraction spells,Brin...%+27788225528 love spells in Toronto Psychic Readings, Attraction spells,Brin...
%+27788225528 love spells in Toronto Psychic Readings, Attraction spells,Brin...
masabamasaba
 
%+27788225528 love spells in Colorado Springs Psychic Readings, Attraction sp...
%+27788225528 love spells in Colorado Springs Psychic Readings, Attraction sp...%+27788225528 love spells in Colorado Springs Psychic Readings, Attraction sp...
%+27788225528 love spells in Colorado Springs Psychic Readings, Attraction sp...
masabamasaba
 

Dernier (20)

WSO2CON 2024 - WSO2's Digital Transformation Journey with Choreo: A Platforml...
WSO2CON 2024 - WSO2's Digital Transformation Journey with Choreo: A Platforml...WSO2CON 2024 - WSO2's Digital Transformation Journey with Choreo: A Platforml...
WSO2CON 2024 - WSO2's Digital Transformation Journey with Choreo: A Platforml...
 
Artyushina_Guest lecture_YorkU CS May 2024.pptx
Artyushina_Guest lecture_YorkU CS May 2024.pptxArtyushina_Guest lecture_YorkU CS May 2024.pptx
Artyushina_Guest lecture_YorkU CS May 2024.pptx
 
%in Midrand+277-882-255-28 abortion pills for sale in midrand
%in Midrand+277-882-255-28 abortion pills for sale in midrand%in Midrand+277-882-255-28 abortion pills for sale in midrand
%in Midrand+277-882-255-28 abortion pills for sale in midrand
 
%+27788225528 love spells in Boston Psychic Readings, Attraction spells,Bring...
%+27788225528 love spells in Boston Psychic Readings, Attraction spells,Bring...%+27788225528 love spells in Boston Psychic Readings, Attraction spells,Bring...
%+27788225528 love spells in Boston Psychic Readings, Attraction spells,Bring...
 
WSO2CON 2024 Slides - Open Source to SaaS
WSO2CON 2024 Slides - Open Source to SaaSWSO2CON 2024 Slides - Open Source to SaaS
WSO2CON 2024 Slides - Open Source to SaaS
 
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
 
%+27788225528 love spells in Knoxville Psychic Readings, Attraction spells,Br...
%+27788225528 love spells in Knoxville Psychic Readings, Attraction spells,Br...%+27788225528 love spells in Knoxville Psychic Readings, Attraction spells,Br...
%+27788225528 love spells in Knoxville Psychic Readings, Attraction spells,Br...
 
OpenChain - The Ramifications of ISO/IEC 5230 and ISO/IEC 18974 for Legal Pro...
OpenChain - The Ramifications of ISO/IEC 5230 and ISO/IEC 18974 for Legal Pro...OpenChain - The Ramifications of ISO/IEC 5230 and ISO/IEC 18974 for Legal Pro...
OpenChain - The Ramifications of ISO/IEC 5230 and ISO/IEC 18974 for Legal Pro...
 
Abortion Pill Prices Boksburg [(+27832195400*)] 🏥 Women's Abortion Clinic in ...
Abortion Pill Prices Boksburg [(+27832195400*)] 🏥 Women's Abortion Clinic in ...Abortion Pill Prices Boksburg [(+27832195400*)] 🏥 Women's Abortion Clinic in ...
Abortion Pill Prices Boksburg [(+27832195400*)] 🏥 Women's Abortion Clinic in ...
 
%in Soweto+277-882-255-28 abortion pills for sale in soweto
%in Soweto+277-882-255-28 abortion pills for sale in soweto%in Soweto+277-882-255-28 abortion pills for sale in soweto
%in Soweto+277-882-255-28 abortion pills for sale in soweto
 
WSO2CON 2024 - API Management Usage at La Poste and Its Impact on Business an...
WSO2CON 2024 - API Management Usage at La Poste and Its Impact on Business an...WSO2CON 2024 - API Management Usage at La Poste and Its Impact on Business an...
WSO2CON 2024 - API Management Usage at La Poste and Its Impact on Business an...
 
WSO2CON 2024 - How to Run a Security Program
WSO2CON 2024 - How to Run a Security ProgramWSO2CON 2024 - How to Run a Security Program
WSO2CON 2024 - How to Run a Security Program
 
WSO2CON 2024 - Navigating API Complexity: REST, GraphQL, gRPC, Websocket, Web...
WSO2CON 2024 - Navigating API Complexity: REST, GraphQL, gRPC, Websocket, Web...WSO2CON 2024 - Navigating API Complexity: REST, GraphQL, gRPC, Websocket, Web...
WSO2CON 2024 - Navigating API Complexity: REST, GraphQL, gRPC, Websocket, Web...
 
%+27788225528 love spells in Toronto Psychic Readings, Attraction spells,Brin...
%+27788225528 love spells in Toronto Psychic Readings, Attraction spells,Brin...%+27788225528 love spells in Toronto Psychic Readings, Attraction spells,Brin...
%+27788225528 love spells in Toronto Psychic Readings, Attraction spells,Brin...
 
Direct Style Effect Systems - The Print[A] Example - A Comprehension Aid
Direct Style Effect Systems -The Print[A] Example- A Comprehension AidDirect Style Effect Systems -The Print[A] Example- A Comprehension Aid
Direct Style Effect Systems - The Print[A] Example - A Comprehension Aid
 
MarTech Trend 2024 Book : Marketing Technology Trends (2024 Edition) How Data...
MarTech Trend 2024 Book : Marketing Technology Trends (2024 Edition) How Data...MarTech Trend 2024 Book : Marketing Technology Trends (2024 Edition) How Data...
MarTech Trend 2024 Book : Marketing Technology Trends (2024 Edition) How Data...
 
WSO2CON 2024 Slides - Unlocking Value with AI
WSO2CON 2024 Slides - Unlocking Value with AIWSO2CON 2024 Slides - Unlocking Value with AI
WSO2CON 2024 Slides - Unlocking Value with AI
 
%+27788225528 love spells in Colorado Springs Psychic Readings, Attraction sp...
%+27788225528 love spells in Colorado Springs Psychic Readings, Attraction sp...%+27788225528 love spells in Colorado Springs Psychic Readings, Attraction sp...
%+27788225528 love spells in Colorado Springs Psychic Readings, Attraction sp...
 
WSO2CON 2024 - Freedom First—Unleashing Developer Potential with Open Source
WSO2CON 2024 - Freedom First—Unleashing Developer Potential with Open SourceWSO2CON 2024 - Freedom First—Unleashing Developer Potential with Open Source
WSO2CON 2024 - Freedom First—Unleashing Developer Potential with Open Source
 
BUS PASS MANGEMENT SYSTEM USING PHP.pptx
BUS PASS MANGEMENT SYSTEM USING PHP.pptxBUS PASS MANGEMENT SYSTEM USING PHP.pptx
BUS PASS MANGEMENT SYSTEM USING PHP.pptx
 

Interactive Tradeoffs Between Competing Offline Metrics with Bayesian Optimization

  • 1. SigOpt. Confidential. Interactive Tradeoffs Between Competing Offline Metrics with Bayesian Optimization KDD 2019 2nd Annual Workshop Online and Offline Evaluation of Interactive Systems Michael McCourt, Research Engineer, SigOpt
  • 2. SigOpt. Confidential. About me ● Research engineering at SigOpt ● Focus on applied Bayesian optimization ● PhD from Cornell ● Avid Cleveland Cavaliers fan About SigOpt ● Leading software solution for parameter optimization and model experimentation ● Customers in finance, trading, media, technology, consulting, energy, industry ● Free version of our solution for academia available at sigopt.com/edu
  • 3. SigOpt. Confidential. Abstract for KDD 2019 2nd Annual Workshop on Offline and Online Evaluation of Interactive Systems Many real world applications (ML models, simulators, etc.) have multiple competing metrics that define performance; these require practitioners to carefully consider potential tradeoffs. However, assessing and ranking this tradeoff is nontrivial, especially when the number of metrics is more than two. Often times, practitioners scalarize the metrics into a single objective, e.g., using a weighted sum. In this talk, we pose this problem as a constrained multi-objective optimization problem. By setting and updating the constraints, we can efficiently explore only the region of the Pareto efficient frontier of the model/system of most interest. We motivate this problem with the application of an experimental design setting, where we are trying to fabricate high performance glass substrate for solar cell panels. 3
  • 4. SigOpt. Confidential. Most Metrics are Impacted by Free Parameters How can these free parameters be chosen? Generally, these are chosen to yield good future performance. • This discussion only covers offline metrics. • Some of the elements apply in an online setting as well. Given a computable metric defining future performance, a search can be conducted for the free parameters yielding acceptable/optimal performance. • In many circumstances, evaluating this performance metric is costly. • Example: Train a classification model and evaluate a validation accuracy. • Example: Use financial data from the past year for a trading strategy and evaluate its profit on last month’s data. 4
  • 5. SigOpt. Confidential. Searching for Free Parameters Requires Efficiency Intelligently searching a fixed domain Many searches benefit from efficiently (actively) learning about the circumstances of the search. • Active learning -- “Active learning is closely related to experimental design … is most often adaptive … employs an oracle for data labelling … is usually used to learn a model for classification.” -- [Brochu et al 2010] Two adjacent fields of research have evolved. • Bayesian optimization -- “Bayesian optimization is a sequential model-based approach to [optimizing a function].” -- [Shahriari et al, 2016] • Active search -- “Active search is an active learning setting with the goal of identifying as many members of a given class as possible under a labeling budget.” -- [Jiang et al, 2017] How we conduct this active learning will greatly impact efficiency of the search. 5
  • 6. SigOpt. Confidential. Bayesian Optimization A graphical depiction of the iterative process 6 Build a statistical model Build a statistical model Choose a next point Choose a next point
  • 7. SigOpt. Confidential. Bayesian Optimization Efficiently Optimize a Scalar Function To quote [Frazier 2018]: Bayesian optimization (BayesOpt/BO) is a class of machine-learning-based optimization methods focused on [maximizing/minimizing a function with] the following properties ... • Typically the dimension d is less than 20. • The objective function f is continuous, as is the domain (which is likely a d-dimensional rectangle). • f is expensive to evaluate: e.g., time, money, access • f is black-box: if lacks known special structure like concavity or linearity. • When we evaluate f, we observe only f(x); that is, the optimization is gradient-free. • f is often observed in the presence of noise. • Our focus is on finding a global rather than local optimum. 7
  • 8. SigOpt. Confidential. Many Metrics may Contribute to Success How do we execute under these ambiguous circumstances? Defining/measuring future performance is imprecise. • Generally, several metrics will contribute to a sense of future performance. • Not all metrics are equally important -- some may only need to reach a threshold. • The feasible performance and preferred interaction between metrics may not be known a priori. Popular multiobjective optimization strategies are often population-based (and not often sample-efficient). 8
  • 9. SigOpt. Confidential. Multiobjective Bayesian Optimization Balancing competing metrics to find the Pareto frontier Adaptations to BO to search for the efficient frontier: • Change the problem to an active search problem [Jiang et al, 2018]. • Search for diverse points near the efficient frontier. • Scalarize the problem with linear combinations of the metrics [Knowles, 2006]. • Define a hypervolume based acquisition function [Hernandez-Lobato et al, 2016, Emmerich et al, 2016]. • Scalarization through prior beliefs [Astudillo, 2017]. Guiding points • Users wanted to interactively update the search process. • Users felt uncomfortable stating a priori preferences. 9
  • 10. SigOpt. Confidential. Multiobjective Bayesian Optimization Our strategy We apply a strategy similar to what was discussed in [Letham et al, 2019]. 1. Model all metrics independently. • Requires no prior beliefs on how metrics interact. • Missing data removed on a per metric basis if unrecorded. 2. Expose the efficient frontier through constrained scalar optimization. • Enforce user constraints when given. • Iterate through sub constraints to better resolve efficient frontier, if desired. • Consider different regions of the frontier when parallelism is possible. 3. Allow users to change constraints as the search progresses. • Allow the problems/goals to evolve as the user’s understanding evolves. 10 Variation on Expected Improvement
  • 11. SigOpt. Confidential. One strategy can be to randomly apply constraints. Multiobjective Bayesian Optimization Our strategy 11 Note: There are GIFs that do not show up in this version of the presentation. For a copy that includes them, please email contact@sigopt.com
  • 12. SigOpt. Confidential. Another strategy can be to “walk” up and down the constraint domain. Multiobjective Bayesian Optimization Our strategy 12 Note: There are GIFs that do not show up in this version of the presentation. For a copy that includes them, please email contact@sigopt.com
  • 13. SigOpt. Confidential. It can help to alternate which metric the constraint is imposed on. Multiobjective Bayesian Optimization Our strategy 13 Note: There are GIFs that do not show up in this version of the presentation. For a copy that includes them, please email contact@sigopt.com
  • 14. SigOpt. Confidential. Users can enforce their own bounds to focus on the desired outcome. Multiobjective Bayesian Optimization Our strategy 14 Note: There are GIFs that do not show up in this version of the presentation. For a copy that includes them, please email contact@sigopt.com
  • 15. SigOpt. Confidential. Users can also update their own bounds as the experiment goes on. Multiobjective Bayesian Optimization Our strategy 15 Note: There are GIFs that do not show up in this version of the presentation. For a copy that includes them, please email contact@sigopt.com
  • 16. SigOpt. Confidential. Awesome Applications of Bayesian Optimization Who is using, and can use, BO? 16 ● ML/DL hyperparameter tuning [Snoek et al, 2012; Feurer et al, 2015; Kandasamy et al, 2018] ● Engineering system design [Mockus, 1989; Jones et al, 1998; Forrester et al, 2008] ● Drug design [Negoescu et al, 2011; Frazier and Wang, 2016] ● Material design [Packwood, 2017; Haghanifar et al, 2019] ● Model calibration [Shoemaker et al, 2007; Shi et al, 2013; Letham et al. 2019] ● Reinforcement learning [Lizotte, 2008; Brochu et al, 2010; Martinez-Cantin et al, 2018] There are so many others!
  • 17. SigOpt. Confidential.17 A Joint Collaboration with University of Pittsburgh [Haghanifar et al, 2019] Metrics • Light transmission • Clarity (low haze) • Water resistance Constraints updated on all metrics during the search. Note: There is a video that does not show up in this version of the presentation. For a copy that includes this, please email contact@sigopt.com
  • 18. SigOpt. Confidential. Future Work How can we improve this process? When black-box constraints exist, how can we encourage our search to respect them? • Hallucinate bad function values at points which violate the constraints. • Attenuate the expected improvement by the probability of failure [Gelbart, 2015]. • Model the constraints and average out the noisy behavior [Letham et al, 2019]. • Model the Lagrangian [Picheny et al, 2016]. Question: Exactly how black-box/expensive are these constraints (or the objective)? • We can adapt to expensive constraints but a cheap objective [Gramacy et al, 2106]. Question: Can we help focus on the important region using preferences? • Joint work extending [Astudillo, 2017] with Raul and Peter. 18