SlideShare a Scribd company logo
Confidential and proprietary: Any use of this material without
specific permission of McKinsey & Company is strictly prohibited
Intelligence.
Beautifully engineered.
No more grid search! How to build models effectively
Thomas Huijskens
London Driverless AI meet-up
1st August, 2019
2All content copyright © 2019 QuantumBlack, a McKinsey company
• Typically, we want to optimize some performance metric 𝑓ℳ for a model ℳ, on a hold-out set of data.
• The model ℳ has some hyperparameters 𝜃 that we need to specify, and the performance of ℳ is highly
dependent on the settings of 𝜃.
• Example: For a classification problem, ℳ could be a support vector machine, and the performance
metric 𝑓ℳ could be the out-of-sample AUC.
• Because the performance of the SVM depends on hyperparameters 𝜃, the metric 𝑓ℳ depends on 𝜃 as
well.
Most models have a number of hyperparameters that need to be chosen
a priori..
3All content copyright © 2019 QuantumBlack, a McKinsey company
.. and our goal is to find the values of 𝜃 for which the value of the (out-
of-sample) performance 𝑓ℳ is optimal
𝑓ℳ 𝜃 𝐷)
Learner ℳ
Performance
function 𝑓
𝑓 depends on the hyperparameters 𝜃 of
the model ℳ
𝑓 is conditional on the
data 𝐷
4All content copyright © 2019 QuantumBlack, a McKinsey company
Manual grid search works for low-dimensional problems, but does not
scale well to higher dimensions
Curse of dimensionality
makes this inefficient in
higher dimensional
problems
5All content copyright © 2019 QuantumBlack, a McKinsey company
Random grid search works better in higher dimensions, but..
6All content copyright © 2019 QuantumBlack, a McKinsey company
.. shouldn't previously evaluated hyperparameter values guide us in the
search for the true optimal value?
Low values of L here,
should probably not
focus in this area
High values of L here,
should probably focus
in this area
7All content copyright © 2019 QuantumBlack, a McKinsey company
Advanced hyper-parameter
optimization methods
8All content copyright © 2019 QuantumBlack, a McKinsey company
Advanced HPO methods are only beneficial in certain situations
In advanced optimisation methods, we spend more computational time to figure out what
set of parameters 𝜃 we are evaluating the performance metric 𝑓ℳ 𝜃 for.
As a result, applying these methods only really makes sense if
• The number of hyperparameters is very high (𝜃 is of high dimension); or
• It is computationally very expensive to evaluate 𝑓ℳ(𝜃) for a single point 𝜃.
9All content copyright © 2019 QuantumBlack, a McKinsey company
There are now a wide range of hyper-parameter optimisation methods
HPO
Variant
Multi-
objective
Multi-fidelity
Search
Random / grid
search
Population-
based
Bayesian
optimisation
SMAC3: https://github.com/automl/SMAC3
Hyperopt: http://hyperopt.github.io/hyperopt/
Ax: https://github.com/facebook/Ax
Tune: https://ray.readthedocs.io/en/latest/tune.html
TPOT: https://github.com/EpistasisLab/tpot
Tune: https://ray.readthedocs.io/en/latest/tune.html
Hyperband: https://github.com/thuijskens/scikit-hyperband
Tune: https://ray.readthedocs.io/en/latest/tune.html
Ax: https://github.com/facebook/Ax
HpBandSter: https://github.com/automl/HpBandSter
Fabolas: https://github.com/automl/RoBO
MTBO: https://github.com/automl/RoBO
10All content copyright © 2019 QuantumBlack, a McKinsey company
BO-HB is a recent method that combines Bayesian optimisation and
multi-fidelity methods
HPO
Variant
Multi-
objective
Multi-fidelity
Search
Random / grid
search
Population-
based
Bayesian
optimisation
SMAC3: https://github.com/automl/SMAC3
Hyperopt: http://hyperopt.github.io/hyperopt/
Ax: https://github.com/facebook/Ax
Tune: https://ray.readthedocs.io/en/latest/tune.html
TPOT: https://github.com/EpistasisLab/tpot
Tune: https://ray.readthedocs.io/en/latest/tune.html
Hyperband: https://github.com/thuijskens/scikit-hyperband
Tune: https://ray.readthedocs.io/en/latest/tune.html
Ax: https://github.com/facebook/Ax
HpBandSter: https://github.com/automl/HpBandSter
Fabolas: https://github.com/automl/RoBO
MTBO: https://github.com/automl/RoBO
11All content copyright © 2019 QuantumBlack, a McKinsey company
Hyperband
12All content copyright © 2019 QuantumBlack, a McKinsey company
Hyperband iteratively runs the successive halving subroutine on model
configurations
BOHB: Robust and efficient hyperparameter optimization at scale: https://www.automl.org/blog_bohb/
13All content copyright © 2019 QuantumBlack, a McKinsey company
Successive halving suffers from the exploitation versus exploration
trade-off
Given a total computational budget, the user has to decide whether
• To try many configurations with a small budget each; or
• To try only a few configurations with a larger budget.
Hyperband hedges for this trade-off by:
• Dividing the total computational budget into several combinations of number of configurations vs.
budget.
• Call successive halving as a subroutine for each combination.
14All content copyright © 2019 QuantumBlack, a McKinsey company
Computational budget
Number of
configurations
Successive halving
iteration
Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization, Lisha Li, Kevin Jamieson, Giulia DeSalvo, Afshin Rostamizadeh, Ameet Talwalkar, 2018. https://arxiv.org/abs/1603.06560
15All content copyright © 2019 QuantumBlack, a McKinsey company
Hyperband’s hedging strategy is very competitive against random
search
BOHB: Robust and Efficient Hyperparameter Optimization at Scale, Stefan Falkner, Aaron Klein, Frank Hutter, 2018. https://arxiv.org/abs/1807.01774
16All content copyright © 2019 QuantumBlack, a McKinsey company
Bayesian optimisation
17All content copyright © 2019 QuantumBlack, a McKinsey company
The classic Bayesian optimization algorithm consists of three steps
Evaluate performance
of learner f with
parameters 𝜃
Update current belief
of loss surface of the
learner f
Choose 𝜃 that
maximises some
utility over the
current belief
𝑓 | 𝑦)*+
𝑦)*+ = 𝑓 𝜃)*+
𝜃)*+
18All content copyright © 2019 QuantumBlack, a McKinsey company
The performance function 𝑓ℳ is modelled to form a current belief of
what the loss surface looks like..
Evaluate performance
of learner f with
parameters 𝜃
Update current belief
of loss surface of the
learner f
Choose 𝜃 that
maximises some
utility over the
current belief
𝑓 | 𝑦)*+
𝑦)*+ = 𝑓 𝜃)*+
𝜃)*+
19All content copyright © 2019 QuantumBlack, a McKinsey company
..acquisition functions are used to formalize what constitutes a ”best
guess”..
Evaluate performance
of learner f with
parameters 𝜃
Update current belief
of loss surface of the
learner f
Choose 𝜃 that
maximises some
utility over the
current belief
𝑓 | 𝑦)*+
𝑦)*+ = 𝑓 𝜃)*+
𝜃)*+
20All content copyright © 2019 QuantumBlack, a McKinsey company
The expected improvement acquisition trades off exploitation of known
optimal areas, versus exploration of unexplored areas of the loss
surface
Exploitation of
known optimal
space
Exploration of
unknown space
A Tutorial on Bayesian Optimization of Expensive Cost Functions, with Application to Active User Modeling and Hierarchical Reinforcement Learning, Eric Brochu, Vlad M. Cora, Nando de Freitas, 2010. https://arxiv.org/abs/1012.2599
21All content copyright © 2019 QuantumBlack, a McKinsey company
.. and finally the performance is computed for suggested parameters
Evaluate performance
of learner f with
parameters 𝜃
Update current belief
of loss surface of the
learner f
Choose 𝜃 that
maximises some
utility over the
current belief
𝑓 | 𝑦)*+
𝑦)*+ = 𝑓 𝜃)*+
𝜃)*+
22All content copyright © 2019 QuantumBlack, a McKinsey company
Bayesian optimization is competitive against random search for longer
training budgets
BOHB: Robust and Efficient Hyperparameter Optimization at Scale, Stefan Falkner, Aaron Klein, Frank Hutter, 2018. https://arxiv.org/abs/1807.01774
23All content copyright © 2019 QuantumBlack, a McKinsey company
Bayesian optimisation works well for larger budgets, whereas
hyperband works better for small to medium budgets
BOHB: Robust and Efficient Hyperparameter Optimization at Scale, Stefan Falkner, Aaron Klein, Frank Hutter, 2018. https://arxiv.org/abs/1807.01774
24All content copyright © 2019 QuantumBlack, a McKinsey company
BO-HB replaces random search in Hyperband with Bayesian
optimisation
BOHB: Robust and Efficient Hyperparameter Optimization at Scale, Stefan Falkner, Aaron Klein, Frank Hutter, 2018. https://arxiv.org/abs/1807.01774
25All content copyright © 2019 QuantumBlack, a McKinsey company
What HPO method should I use in practice?
• If it is possible to define substantially cheaper version of the performance function 𝑓ℳ:
Bayesian optimisation-hyperband (BO-HB)
• If this is not possible and
• if all hyperparameters are real-valued and one can only afford a few dozen function
evaluations: Gaussian-process based Bayesian optimisation (SMAC, Spearmint)
• if the configuration space is large and/or conditional: Random forest-based Bayesian
optimisation (hyperopt)
Questions?

More Related Content

What's hot

Nick Schmidt, BLDS - Responsible Data Science: Identifying and Fixing Biased ...
Nick Schmidt, BLDS - Responsible Data Science: Identifying and Fixing Biased ...Nick Schmidt, BLDS - Responsible Data Science: Identifying and Fixing Biased ...
Nick Schmidt, BLDS - Responsible Data Science: Identifying and Fixing Biased ...Sri Ambati
 
Tom Aliff, Equifax - Configurable Modeling for Maximizing Business Value - H2...
Tom Aliff, Equifax - Configurable Modeling for Maximizing Business Value - H2...Tom Aliff, Equifax - Configurable Modeling for Maximizing Business Value - H2...
Tom Aliff, Equifax - Configurable Modeling for Maximizing Business Value - H2...Sri Ambati
 
Introduction & Hands-on with H2O Driverless AI
Introduction & Hands-on with H2O Driverless AIIntroduction & Hands-on with H2O Driverless AI
Introduction & Hands-on with H2O Driverless AISri Ambati
 
Rahul Bhuman, Tech Mahindra - Truck roll prediction using Driverless AI - H2O...
Rahul Bhuman, Tech Mahindra - Truck roll prediction using Driverless AI - H2O...Rahul Bhuman, Tech Mahindra - Truck roll prediction using Driverless AI - H2O...
Rahul Bhuman, Tech Mahindra - Truck roll prediction using Driverless AI - H2O...Sri Ambati
 
Mark Seiss, Dun & Bradstreet - Importance of Domain Expertise for Building ML...
Mark Seiss, Dun & Bradstreet - Importance of Domain Expertise for Building ML...Mark Seiss, Dun & Bradstreet - Importance of Domain Expertise for Building ML...
Mark Seiss, Dun & Bradstreet - Importance of Domain Expertise for Building ML...Sri Ambati
 
Towards Human-Centered Machine Learning
Towards Human-Centered Machine LearningTowards Human-Centered Machine Learning
Towards Human-Centered Machine LearningSri Ambati
 
Carmelo Iaria, AI Academy - How The AI Academy is accelerating NLP projects w...
Carmelo Iaria, AI Academy - How The AI Academy is accelerating NLP projects w...Carmelo Iaria, AI Academy - How The AI Academy is accelerating NLP projects w...
Carmelo Iaria, AI Academy - How The AI Academy is accelerating NLP projects w...Sri Ambati
 
TechKnow Fiesta 2021 - Powered by Amazon Web Service in September 2021
TechKnow Fiesta 2021 - Powered by Amazon Web Service in September 2021TechKnow Fiesta 2021 - Powered by Amazon Web Service in September 2021
TechKnow Fiesta 2021 - Powered by Amazon Web Service in September 2021TechknowFiesta
 
Democratization - New Wave of Data Science (홍운표 상무, DataRobot) :: AWS Techfor...
Democratization - New Wave of Data Science (홍운표 상무, DataRobot) :: AWS Techfor...Democratization - New Wave of Data Science (홍운표 상무, DataRobot) :: AWS Techfor...
Democratization - New Wave of Data Science (홍운표 상무, DataRobot) :: AWS Techfor...Amazon Web Services Korea
 
Dive into H2O: NYC
Dive into H2O: NYCDive into H2O: NYC
Dive into H2O: NYCSri Ambati
 
Data science with python certification training course with
Data science with python certification training course withData science with python certification training course with
Data science with python certification training course withkiruthikab6
 
Enterprise deep learning lessons bodkin o reilly ai sf 2017
Enterprise deep learning lessons bodkin o reilly ai sf 2017Enterprise deep learning lessons bodkin o reilly ai sf 2017
Enterprise deep learning lessons bodkin o reilly ai sf 2017Ron Bodkin
 
Graph + AI World 2020: Opening Day Keynote
Graph + AI World 2020: Opening Day KeynoteGraph + AI World 2020: Opening Day Keynote
Graph + AI World 2020: Opening Day KeynoteTigerGraph
 
Robert Coop, Stanley Black & Decker - Optimizing Manufacturing with Driverles...
Robert Coop, Stanley Black & Decker - Optimizing Manufacturing with Driverles...Robert Coop, Stanley Black & Decker - Optimizing Manufacturing with Driverles...
Robert Coop, Stanley Black & Decker - Optimizing Manufacturing with Driverles...Sri Ambati
 
Predicting Consumer Behaviour via Hadoop
Predicting Consumer Behaviour via HadoopPredicting Consumer Behaviour via Hadoop
Predicting Consumer Behaviour via HadoopSkillspeed
 
AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...
AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...
AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...Sri Ambati
 
Introduction to pyspark new
Introduction to pyspark newIntroduction to pyspark new
Introduction to pyspark newAnam Mahmood
 
AI in the Enterprise
AI in the EnterpriseAI in the Enterprise
AI in the EnterpriseRon Bodkin
 
Patrick Hall, H2O.ai - Human Friendly Machine Learning - H2O World San Francisco
Patrick Hall, H2O.ai - Human Friendly Machine Learning - H2O World San FranciscoPatrick Hall, H2O.ai - Human Friendly Machine Learning - H2O World San Francisco
Patrick Hall, H2O.ai - Human Friendly Machine Learning - H2O World San FranciscoSri Ambati
 
Datasciencein E-commerce industry
Datasciencein E-commerce industryDatasciencein E-commerce industry
Datasciencein E-commerce industryRakuten Group, Inc.
 

What's hot (20)

Nick Schmidt, BLDS - Responsible Data Science: Identifying and Fixing Biased ...
Nick Schmidt, BLDS - Responsible Data Science: Identifying and Fixing Biased ...Nick Schmidt, BLDS - Responsible Data Science: Identifying and Fixing Biased ...
Nick Schmidt, BLDS - Responsible Data Science: Identifying and Fixing Biased ...
 
Tom Aliff, Equifax - Configurable Modeling for Maximizing Business Value - H2...
Tom Aliff, Equifax - Configurable Modeling for Maximizing Business Value - H2...Tom Aliff, Equifax - Configurable Modeling for Maximizing Business Value - H2...
Tom Aliff, Equifax - Configurable Modeling for Maximizing Business Value - H2...
 
Introduction & Hands-on with H2O Driverless AI
Introduction & Hands-on with H2O Driverless AIIntroduction & Hands-on with H2O Driverless AI
Introduction & Hands-on with H2O Driverless AI
 
Rahul Bhuman, Tech Mahindra - Truck roll prediction using Driverless AI - H2O...
Rahul Bhuman, Tech Mahindra - Truck roll prediction using Driverless AI - H2O...Rahul Bhuman, Tech Mahindra - Truck roll prediction using Driverless AI - H2O...
Rahul Bhuman, Tech Mahindra - Truck roll prediction using Driverless AI - H2O...
 
Mark Seiss, Dun & Bradstreet - Importance of Domain Expertise for Building ML...
Mark Seiss, Dun & Bradstreet - Importance of Domain Expertise for Building ML...Mark Seiss, Dun & Bradstreet - Importance of Domain Expertise for Building ML...
Mark Seiss, Dun & Bradstreet - Importance of Domain Expertise for Building ML...
 
Towards Human-Centered Machine Learning
Towards Human-Centered Machine LearningTowards Human-Centered Machine Learning
Towards Human-Centered Machine Learning
 
Carmelo Iaria, AI Academy - How The AI Academy is accelerating NLP projects w...
Carmelo Iaria, AI Academy - How The AI Academy is accelerating NLP projects w...Carmelo Iaria, AI Academy - How The AI Academy is accelerating NLP projects w...
Carmelo Iaria, AI Academy - How The AI Academy is accelerating NLP projects w...
 
TechKnow Fiesta 2021 - Powered by Amazon Web Service in September 2021
TechKnow Fiesta 2021 - Powered by Amazon Web Service in September 2021TechKnow Fiesta 2021 - Powered by Amazon Web Service in September 2021
TechKnow Fiesta 2021 - Powered by Amazon Web Service in September 2021
 
Democratization - New Wave of Data Science (홍운표 상무, DataRobot) :: AWS Techfor...
Democratization - New Wave of Data Science (홍운표 상무, DataRobot) :: AWS Techfor...Democratization - New Wave of Data Science (홍운표 상무, DataRobot) :: AWS Techfor...
Democratization - New Wave of Data Science (홍운표 상무, DataRobot) :: AWS Techfor...
 
Dive into H2O: NYC
Dive into H2O: NYCDive into H2O: NYC
Dive into H2O: NYC
 
Data science with python certification training course with
Data science with python certification training course withData science with python certification training course with
Data science with python certification training course with
 
Enterprise deep learning lessons bodkin o reilly ai sf 2017
Enterprise deep learning lessons bodkin o reilly ai sf 2017Enterprise deep learning lessons bodkin o reilly ai sf 2017
Enterprise deep learning lessons bodkin o reilly ai sf 2017
 
Graph + AI World 2020: Opening Day Keynote
Graph + AI World 2020: Opening Day KeynoteGraph + AI World 2020: Opening Day Keynote
Graph + AI World 2020: Opening Day Keynote
 
Robert Coop, Stanley Black & Decker - Optimizing Manufacturing with Driverles...
Robert Coop, Stanley Black & Decker - Optimizing Manufacturing with Driverles...Robert Coop, Stanley Black & Decker - Optimizing Manufacturing with Driverles...
Robert Coop, Stanley Black & Decker - Optimizing Manufacturing with Driverles...
 
Predicting Consumer Behaviour via Hadoop
Predicting Consumer Behaviour via HadoopPredicting Consumer Behaviour via Hadoop
Predicting Consumer Behaviour via Hadoop
 
AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...
AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...
AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transfo...
 
Introduction to pyspark new
Introduction to pyspark newIntroduction to pyspark new
Introduction to pyspark new
 
AI in the Enterprise
AI in the EnterpriseAI in the Enterprise
AI in the Enterprise
 
Patrick Hall, H2O.ai - Human Friendly Machine Learning - H2O World San Francisco
Patrick Hall, H2O.ai - Human Friendly Machine Learning - H2O World San FranciscoPatrick Hall, H2O.ai - Human Friendly Machine Learning - H2O World San Francisco
Patrick Hall, H2O.ai - Human Friendly Machine Learning - H2O World San Francisco
 
Datasciencein E-commerce industry
Datasciencein E-commerce industryDatasciencein E-commerce industry
Datasciencein E-commerce industry
 

Similar to No more grid search! How to build models effectively by Thomas Huijskens

Cloud Deep Dive: Total Cost of Ownership - John Enoch
Cloud Deep Dive: Total Cost of Ownership - John EnochCloud Deep Dive: Total Cost of Ownership - John Enoch
Cloud Deep Dive: Total Cost of Ownership - John EnochAmazon Web Services
 
PyData London 2018 talk on feature selection
PyData London 2018 talk on feature selectionPyData London 2018 talk on feature selection
PyData London 2018 talk on feature selectionThomas Huijskens
 
Mainframe MRI from CA Technologies
Mainframe MRI from CA TechnologiesMainframe MRI from CA Technologies
Mainframe MRI from CA TechnologiesCA Technologies
 
Product Development in the Cloud
Product Development in the Cloud Product Development in the Cloud
Product Development in the Cloud Amazon Web Services
 
Increase revenue with header bidding | adapex
Increase revenue with header bidding | adapexIncrease revenue with header bidding | adapex
Increase revenue with header bidding | adapexAdapex LLC
 
Transforming Product Development in the Cloud (ENT306) - AWS re:Invent 2018
Transforming Product Development in the Cloud (ENT306) - AWS re:Invent 2018Transforming Product Development in the Cloud (ENT306) - AWS re:Invent 2018
Transforming Product Development in the Cloud (ENT306) - AWS re:Invent 2018Amazon Web Services
 
Product Development in the Cloud - ENT206 - Chicago AWS Summit
Product Development in the Cloud - ENT206 - Chicago AWS SummitProduct Development in the Cloud - ENT206 - Chicago AWS Summit
Product Development in the Cloud - ENT206 - Chicago AWS SummitAmazon Web Services
 
ENT206 Product Development in the Cloud
ENT206 Product Development in the CloudENT206 Product Development in the Cloud
ENT206 Product Development in the CloudAmazon Web Services
 
AWS Enterprise Summit Netherlands - Cost Optimisation at Scale
AWS Enterprise Summit Netherlands - Cost Optimisation at ScaleAWS Enterprise Summit Netherlands - Cost Optimisation at Scale
AWS Enterprise Summit Netherlands - Cost Optimisation at ScaleAmazon Web Services
 
Driving Business Goals with Recommender Systems @ YAC/m 2015
Driving Business Goals with Recommender Systems @ YAC/m 2015Driving Business Goals with Recommender Systems @ YAC/m 2015
Driving Business Goals with Recommender Systems @ YAC/m 2015Konstantin Savenkov
 
Organizing for faster innovation - People, process, culture, and technology
Organizing for faster innovation - People, process, culture, and technologyOrganizing for faster innovation - People, process, culture, and technology
Organizing for faster innovation - People, process, culture, and technologyTom Laszewski
 
Monolithic to Heterogenous - AWS FS Cloud Symposium Apr 2019.pdf
Monolithic to Heterogenous - AWS FS Cloud Symposium Apr 2019.pdfMonolithic to Heterogenous - AWS FS Cloud Symposium Apr 2019.pdf
Monolithic to Heterogenous - AWS FS Cloud Symposium Apr 2019.pdfAmazon Web Services
 
Pricing Optimization using Machine Learning
Pricing Optimization using Machine LearningPricing Optimization using Machine Learning
Pricing Optimization using Machine LearningIRJET Journal
 
Transforming Product Development - Transformation Day Montreal 2018
Transforming Product Development - Transformation Day Montreal 2018Transforming Product Development - Transformation Day Montreal 2018
Transforming Product Development - Transformation Day Montreal 2018Amazon Web Services
 
Cost Optimization for Apache Hadoop/Spark Workloads with Amazon EMR
Cost Optimization for Apache Hadoop/Spark Workloads with Amazon EMRCost Optimization for Apache Hadoop/Spark Workloads with Amazon EMR
Cost Optimization for Apache Hadoop/Spark Workloads with Amazon EMRProvectus
 
churn_detection.pptx
churn_detection.pptxchurn_detection.pptx
churn_detection.pptxDhanuDhanu49
 

Similar to No more grid search! How to build models effectively by Thomas Huijskens (20)

2019 Triangle Machine Learning Day - Stacking Audience Models -- Using an Ens...
2019 Triangle Machine Learning Day - Stacking Audience Models -- Using an Ens...2019 Triangle Machine Learning Day - Stacking Audience Models -- Using an Ens...
2019 Triangle Machine Learning Day - Stacking Audience Models -- Using an Ens...
 
Cloud Deep Dive: Total Cost of Ownership - John Enoch
Cloud Deep Dive: Total Cost of Ownership - John EnochCloud Deep Dive: Total Cost of Ownership - John Enoch
Cloud Deep Dive: Total Cost of Ownership - John Enoch
 
Pydata presentation
Pydata presentationPydata presentation
Pydata presentation
 
PyData London 2018 talk on feature selection
PyData London 2018 talk on feature selectionPyData London 2018 talk on feature selection
PyData London 2018 talk on feature selection
 
Mainframe MRI from CA Technologies
Mainframe MRI from CA TechnologiesMainframe MRI from CA Technologies
Mainframe MRI from CA Technologies
 
Product Development in the Cloud
Product Development in the Cloud Product Development in the Cloud
Product Development in the Cloud
 
Cloud Economics - TCO 101
Cloud Economics - TCO 101Cloud Economics - TCO 101
Cloud Economics - TCO 101
 
Increase revenue with header bidding | adapex
Increase revenue with header bidding | adapexIncrease revenue with header bidding | adapex
Increase revenue with header bidding | adapex
 
Transforming Product Development in the Cloud (ENT306) - AWS re:Invent 2018
Transforming Product Development in the Cloud (ENT306) - AWS re:Invent 2018Transforming Product Development in the Cloud (ENT306) - AWS re:Invent 2018
Transforming Product Development in the Cloud (ENT306) - AWS re:Invent 2018
 
Product Development in the Cloud - ENT206 - Chicago AWS Summit
Product Development in the Cloud - ENT206 - Chicago AWS SummitProduct Development in the Cloud - ENT206 - Chicago AWS Summit
Product Development in the Cloud - ENT206 - Chicago AWS Summit
 
ENT206 Product Development in the Cloud
ENT206 Product Development in the CloudENT206 Product Development in the Cloud
ENT206 Product Development in the Cloud
 
AWS Enterprise Summit Netherlands - Cost Optimisation at Scale
AWS Enterprise Summit Netherlands - Cost Optimisation at ScaleAWS Enterprise Summit Netherlands - Cost Optimisation at Scale
AWS Enterprise Summit Netherlands - Cost Optimisation at Scale
 
Cost Optimisation on AWS
Cost Optimisation on AWSCost Optimisation on AWS
Cost Optimisation on AWS
 
Driving Business Goals with Recommender Systems @ YAC/m 2015
Driving Business Goals with Recommender Systems @ YAC/m 2015Driving Business Goals with Recommender Systems @ YAC/m 2015
Driving Business Goals with Recommender Systems @ YAC/m 2015
 
Organizing for faster innovation - People, process, culture, and technology
Organizing for faster innovation - People, process, culture, and technologyOrganizing for faster innovation - People, process, culture, and technology
Organizing for faster innovation - People, process, culture, and technology
 
Monolithic to Heterogenous - AWS FS Cloud Symposium Apr 2019.pdf
Monolithic to Heterogenous - AWS FS Cloud Symposium Apr 2019.pdfMonolithic to Heterogenous - AWS FS Cloud Symposium Apr 2019.pdf
Monolithic to Heterogenous - AWS FS Cloud Symposium Apr 2019.pdf
 
Pricing Optimization using Machine Learning
Pricing Optimization using Machine LearningPricing Optimization using Machine Learning
Pricing Optimization using Machine Learning
 
Transforming Product Development - Transformation Day Montreal 2018
Transforming Product Development - Transformation Day Montreal 2018Transforming Product Development - Transformation Day Montreal 2018
Transforming Product Development - Transformation Day Montreal 2018
 
Cost Optimization for Apache Hadoop/Spark Workloads with Amazon EMR
Cost Optimization for Apache Hadoop/Spark Workloads with Amazon EMRCost Optimization for Apache Hadoop/Spark Workloads with Amazon EMR
Cost Optimization for Apache Hadoop/Spark Workloads with Amazon EMR
 
churn_detection.pptx
churn_detection.pptxchurn_detection.pptx
churn_detection.pptx
 

More from Sri Ambati

GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...Sri Ambati
 
H2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo Day
H2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo DayH2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo Day
H2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo DaySri Ambati
 
Generative AI Masterclass - Model Risk Management.pptx
Generative AI Masterclass - Model Risk Management.pptxGenerative AI Masterclass - Model Risk Management.pptx
Generative AI Masterclass - Model Risk Management.pptxSri Ambati
 
AI and the Future of Software Development: A Sneak Peek
AI and the Future of Software Development: A Sneak Peek AI and the Future of Software Development: A Sneak Peek
AI and the Future of Software Development: A Sneak Peek Sri Ambati
 
LLMOps: Match report from the top of the 5th
LLMOps: Match report from the top of the 5thLLMOps: Match report from the top of the 5th
LLMOps: Match report from the top of the 5thSri Ambati
 
Building, Evaluating, and Optimizing your RAG App for Production
Building, Evaluating, and Optimizing your RAG App for ProductionBuilding, Evaluating, and Optimizing your RAG App for Production
Building, Evaluating, and Optimizing your RAG App for ProductionSri Ambati
 
Building LLM Solutions using Open Source and Closed Source Solutions in Coher...
Building LLM Solutions using Open Source and Closed Source Solutions in Coher...Building LLM Solutions using Open Source and Closed Source Solutions in Coher...
Building LLM Solutions using Open Source and Closed Source Solutions in Coher...Sri Ambati
 
Risk Management for LLMs
Risk Management for LLMsRisk Management for LLMs
Risk Management for LLMsSri Ambati
 
Open-Source AI: Community is the Way
Open-Source AI: Community is the WayOpen-Source AI: Community is the Way
Open-Source AI: Community is the WaySri Ambati
 
Building Custom GenAI Apps at H2O
Building Custom GenAI Apps at H2OBuilding Custom GenAI Apps at H2O
Building Custom GenAI Apps at H2OSri Ambati
 
Applied Gen AI for the Finance Vertical
Applied Gen AI for the Finance Vertical Applied Gen AI for the Finance Vertical
Applied Gen AI for the Finance Vertical Sri Ambati
 
Cutting Edge Tricks from LLM Papers
Cutting Edge Tricks from LLM PapersCutting Edge Tricks from LLM Papers
Cutting Edge Tricks from LLM PapersSri Ambati
 
Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...
Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...
Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...Sri Ambati
 
Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...
Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...
Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...Sri Ambati
 
KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...
KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...
KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...Sri Ambati
 
LLM Interpretability
LLM Interpretability LLM Interpretability
LLM Interpretability Sri Ambati
 
Never Reply to an Email Again
Never Reply to an Email AgainNever Reply to an Email Again
Never Reply to an Email AgainSri Ambati
 
Introducción al Aprendizaje Automatico con H2O-3 (1)
Introducción al Aprendizaje Automatico con H2O-3 (1)Introducción al Aprendizaje Automatico con H2O-3 (1)
Introducción al Aprendizaje Automatico con H2O-3 (1)Sri Ambati
 
From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...
From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...
From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...Sri Ambati
 
AI Foundations Course Module 1 - An AI Transformation Journey
AI Foundations Course Module 1 - An AI Transformation JourneyAI Foundations Course Module 1 - An AI Transformation Journey
AI Foundations Course Module 1 - An AI Transformation JourneySri Ambati
 

More from Sri Ambati (20)

GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
GenAISummit 2024 May 28 Sri Ambati Keynote: AGI Belongs to The Community in O...
 
H2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo Day
H2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo DayH2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo Day
H2O.ai CEO/Founder: Sri Ambati Keynote at Wells Fargo Day
 
Generative AI Masterclass - Model Risk Management.pptx
Generative AI Masterclass - Model Risk Management.pptxGenerative AI Masterclass - Model Risk Management.pptx
Generative AI Masterclass - Model Risk Management.pptx
 
AI and the Future of Software Development: A Sneak Peek
AI and the Future of Software Development: A Sneak Peek AI and the Future of Software Development: A Sneak Peek
AI and the Future of Software Development: A Sneak Peek
 
LLMOps: Match report from the top of the 5th
LLMOps: Match report from the top of the 5thLLMOps: Match report from the top of the 5th
LLMOps: Match report from the top of the 5th
 
Building, Evaluating, and Optimizing your RAG App for Production
Building, Evaluating, and Optimizing your RAG App for ProductionBuilding, Evaluating, and Optimizing your RAG App for Production
Building, Evaluating, and Optimizing your RAG App for Production
 
Building LLM Solutions using Open Source and Closed Source Solutions in Coher...
Building LLM Solutions using Open Source and Closed Source Solutions in Coher...Building LLM Solutions using Open Source and Closed Source Solutions in Coher...
Building LLM Solutions using Open Source and Closed Source Solutions in Coher...
 
Risk Management for LLMs
Risk Management for LLMsRisk Management for LLMs
Risk Management for LLMs
 
Open-Source AI: Community is the Way
Open-Source AI: Community is the WayOpen-Source AI: Community is the Way
Open-Source AI: Community is the Way
 
Building Custom GenAI Apps at H2O
Building Custom GenAI Apps at H2OBuilding Custom GenAI Apps at H2O
Building Custom GenAI Apps at H2O
 
Applied Gen AI for the Finance Vertical
Applied Gen AI for the Finance Vertical Applied Gen AI for the Finance Vertical
Applied Gen AI for the Finance Vertical
 
Cutting Edge Tricks from LLM Papers
Cutting Edge Tricks from LLM PapersCutting Edge Tricks from LLM Papers
Cutting Edge Tricks from LLM Papers
 
Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...
Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...
Practitioner's Guide to LLMs: Exploring Use Cases and a Glimpse Beyond Curren...
 
Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...
Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...
Open Source h2oGPT with Retrieval Augmented Generation (RAG), Web Search, and...
 
KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...
KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...
KGM Mastering Classification and Regression with LLMs: Insights from Kaggle C...
 
LLM Interpretability
LLM Interpretability LLM Interpretability
LLM Interpretability
 
Never Reply to an Email Again
Never Reply to an Email AgainNever Reply to an Email Again
Never Reply to an Email Again
 
Introducción al Aprendizaje Automatico con H2O-3 (1)
Introducción al Aprendizaje Automatico con H2O-3 (1)Introducción al Aprendizaje Automatico con H2O-3 (1)
Introducción al Aprendizaje Automatico con H2O-3 (1)
 
From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...
From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...
From Rapid Prototypes to an end-to-end Model Deployment: an AI Hedge Fund Use...
 
AI Foundations Course Module 1 - An AI Transformation Journey
AI Foundations Course Module 1 - An AI Transformation JourneyAI Foundations Course Module 1 - An AI Transformation Journey
AI Foundations Course Module 1 - An AI Transformation Journey
 

Recently uploaded

Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Jeffrey Haguewood
 
From Daily Decisions to Bottom Line: Connecting Product Work to Revenue by VP...
From Daily Decisions to Bottom Line: Connecting Product Work to Revenue by VP...From Daily Decisions to Bottom Line: Connecting Product Work to Revenue by VP...
From Daily Decisions to Bottom Line: Connecting Product Work to Revenue by VP...Product School
 
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMsTo Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMsPaul Groth
 
IOS-PENTESTING-BEGINNERS-PRACTICAL-GUIDE-.pptx
IOS-PENTESTING-BEGINNERS-PRACTICAL-GUIDE-.pptxIOS-PENTESTING-BEGINNERS-PRACTICAL-GUIDE-.pptx
IOS-PENTESTING-BEGINNERS-PRACTICAL-GUIDE-.pptxAbida Shariff
 
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Product School
 
Powerful Start- the Key to Project Success, Barbara Laskowska
Powerful Start- the Key to Project Success, Barbara LaskowskaPowerful Start- the Key to Project Success, Barbara Laskowska
Powerful Start- the Key to Project Success, Barbara LaskowskaCzechDreamin
 
Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality
Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered QualitySoftware Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality
Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered QualityInflectra
 
AI revolution and Salesforce, Jiří Karpíšek
AI revolution and Salesforce, Jiří KarpíšekAI revolution and Salesforce, Jiří Karpíšek
AI revolution and Salesforce, Jiří KarpíšekCzechDreamin
 
Custom Approval Process: A New Perspective, Pavel Hrbacek & Anindya Halder
Custom Approval Process: A New Perspective, Pavel Hrbacek & Anindya HalderCustom Approval Process: A New Perspective, Pavel Hrbacek & Anindya Halder
Custom Approval Process: A New Perspective, Pavel Hrbacek & Anindya HalderCzechDreamin
 
Unsubscribed: Combat Subscription Fatigue With a Membership Mentality by Head...
Unsubscribed: Combat Subscription Fatigue With a Membership Mentality by Head...Unsubscribed: Combat Subscription Fatigue With a Membership Mentality by Head...
Unsubscribed: Combat Subscription Fatigue With a Membership Mentality by Head...Product School
 
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024Tobias Schneck
 
Exploring UiPath Orchestrator API: updates and limits in 2024 🚀
Exploring UiPath Orchestrator API: updates and limits in 2024 🚀Exploring UiPath Orchestrator API: updates and limits in 2024 🚀
Exploring UiPath Orchestrator API: updates and limits in 2024 🚀DianaGray10
 
ODC, Data Fabric and Architecture User Group
ODC, Data Fabric and Architecture User GroupODC, Data Fabric and Architecture User Group
ODC, Data Fabric and Architecture User GroupCatarinaPereira64715
 
In-Depth Performance Testing Guide for IT Professionals
In-Depth Performance Testing Guide for IT ProfessionalsIn-Depth Performance Testing Guide for IT Professionals
In-Depth Performance Testing Guide for IT ProfessionalsExpeed Software
 
Measures in SQL (a talk at SF Distributed Systems meetup, 2024-05-22)
Measures in SQL (a talk at SF Distributed Systems meetup, 2024-05-22)Measures in SQL (a talk at SF Distributed Systems meetup, 2024-05-22)
Measures in SQL (a talk at SF Distributed Systems meetup, 2024-05-22)Julian Hyde
 
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...UiPathCommunity
 
Knowledge engineering: from people to machines and back
Knowledge engineering: from people to machines and backKnowledge engineering: from people to machines and back
Knowledge engineering: from people to machines and backElena Simperl
 
Search and Society: Reimagining Information Access for Radical Futures
Search and Society: Reimagining Information Access for Radical FuturesSearch and Society: Reimagining Information Access for Radical Futures
Search and Society: Reimagining Information Access for Radical FuturesBhaskar Mitra
 
Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...Product School
 
Assuring Contact Center Experiences for Your Customers With ThousandEyes
Assuring Contact Center Experiences for Your Customers With ThousandEyesAssuring Contact Center Experiences for Your Customers With ThousandEyes
Assuring Contact Center Experiences for Your Customers With ThousandEyesThousandEyes
 

Recently uploaded (20)

Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
Slack (or Teams) Automation for Bonterra Impact Management (fka Social Soluti...
 
From Daily Decisions to Bottom Line: Connecting Product Work to Revenue by VP...
From Daily Decisions to Bottom Line: Connecting Product Work to Revenue by VP...From Daily Decisions to Bottom Line: Connecting Product Work to Revenue by VP...
From Daily Decisions to Bottom Line: Connecting Product Work to Revenue by VP...
 
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMsTo Graph or Not to Graph Knowledge Graph Architectures and LLMs
To Graph or Not to Graph Knowledge Graph Architectures and LLMs
 
IOS-PENTESTING-BEGINNERS-PRACTICAL-GUIDE-.pptx
IOS-PENTESTING-BEGINNERS-PRACTICAL-GUIDE-.pptxIOS-PENTESTING-BEGINNERS-PRACTICAL-GUIDE-.pptx
IOS-PENTESTING-BEGINNERS-PRACTICAL-GUIDE-.pptx
 
Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...Mission to Decommission: Importance of Decommissioning Products to Increase E...
Mission to Decommission: Importance of Decommissioning Products to Increase E...
 
Powerful Start- the Key to Project Success, Barbara Laskowska
Powerful Start- the Key to Project Success, Barbara LaskowskaPowerful Start- the Key to Project Success, Barbara Laskowska
Powerful Start- the Key to Project Success, Barbara Laskowska
 
Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality
Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered QualitySoftware Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality
Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality
 
AI revolution and Salesforce, Jiří Karpíšek
AI revolution and Salesforce, Jiří KarpíšekAI revolution and Salesforce, Jiří Karpíšek
AI revolution and Salesforce, Jiří Karpíšek
 
Custom Approval Process: A New Perspective, Pavel Hrbacek & Anindya Halder
Custom Approval Process: A New Perspective, Pavel Hrbacek & Anindya HalderCustom Approval Process: A New Perspective, Pavel Hrbacek & Anindya Halder
Custom Approval Process: A New Perspective, Pavel Hrbacek & Anindya Halder
 
Unsubscribed: Combat Subscription Fatigue With a Membership Mentality by Head...
Unsubscribed: Combat Subscription Fatigue With a Membership Mentality by Head...Unsubscribed: Combat Subscription Fatigue With a Membership Mentality by Head...
Unsubscribed: Combat Subscription Fatigue With a Membership Mentality by Head...
 
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
Kubernetes & AI - Beauty and the Beast !?! @KCD Istanbul 2024
 
Exploring UiPath Orchestrator API: updates and limits in 2024 🚀
Exploring UiPath Orchestrator API: updates and limits in 2024 🚀Exploring UiPath Orchestrator API: updates and limits in 2024 🚀
Exploring UiPath Orchestrator API: updates and limits in 2024 🚀
 
ODC, Data Fabric and Architecture User Group
ODC, Data Fabric and Architecture User GroupODC, Data Fabric and Architecture User Group
ODC, Data Fabric and Architecture User Group
 
In-Depth Performance Testing Guide for IT Professionals
In-Depth Performance Testing Guide for IT ProfessionalsIn-Depth Performance Testing Guide for IT Professionals
In-Depth Performance Testing Guide for IT Professionals
 
Measures in SQL (a talk at SF Distributed Systems meetup, 2024-05-22)
Measures in SQL (a talk at SF Distributed Systems meetup, 2024-05-22)Measures in SQL (a talk at SF Distributed Systems meetup, 2024-05-22)
Measures in SQL (a talk at SF Distributed Systems meetup, 2024-05-22)
 
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...
 
Knowledge engineering: from people to machines and back
Knowledge engineering: from people to machines and backKnowledge engineering: from people to machines and back
Knowledge engineering: from people to machines and back
 
Search and Society: Reimagining Information Access for Radical Futures
Search and Society: Reimagining Information Access for Radical FuturesSearch and Society: Reimagining Information Access for Radical Futures
Search and Society: Reimagining Information Access for Radical Futures
 
Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...Designing Great Products: The Power of Design and Leadership by Chief Designe...
Designing Great Products: The Power of Design and Leadership by Chief Designe...
 
Assuring Contact Center Experiences for Your Customers With ThousandEyes
Assuring Contact Center Experiences for Your Customers With ThousandEyesAssuring Contact Center Experiences for Your Customers With ThousandEyes
Assuring Contact Center Experiences for Your Customers With ThousandEyes
 

No more grid search! How to build models effectively by Thomas Huijskens

  • 1. Confidential and proprietary: Any use of this material without specific permission of McKinsey & Company is strictly prohibited Intelligence. Beautifully engineered. No more grid search! How to build models effectively Thomas Huijskens London Driverless AI meet-up 1st August, 2019
  • 2. 2All content copyright © 2019 QuantumBlack, a McKinsey company • Typically, we want to optimize some performance metric 𝑓ℳ for a model ℳ, on a hold-out set of data. • The model ℳ has some hyperparameters 𝜃 that we need to specify, and the performance of ℳ is highly dependent on the settings of 𝜃. • Example: For a classification problem, ℳ could be a support vector machine, and the performance metric 𝑓ℳ could be the out-of-sample AUC. • Because the performance of the SVM depends on hyperparameters 𝜃, the metric 𝑓ℳ depends on 𝜃 as well. Most models have a number of hyperparameters that need to be chosen a priori..
  • 3. 3All content copyright © 2019 QuantumBlack, a McKinsey company .. and our goal is to find the values of 𝜃 for which the value of the (out- of-sample) performance 𝑓ℳ is optimal 𝑓ℳ 𝜃 𝐷) Learner ℳ Performance function 𝑓 𝑓 depends on the hyperparameters 𝜃 of the model ℳ 𝑓 is conditional on the data 𝐷
  • 4. 4All content copyright © 2019 QuantumBlack, a McKinsey company Manual grid search works for low-dimensional problems, but does not scale well to higher dimensions Curse of dimensionality makes this inefficient in higher dimensional problems
  • 5. 5All content copyright © 2019 QuantumBlack, a McKinsey company Random grid search works better in higher dimensions, but..
  • 6. 6All content copyright © 2019 QuantumBlack, a McKinsey company .. shouldn't previously evaluated hyperparameter values guide us in the search for the true optimal value? Low values of L here, should probably not focus in this area High values of L here, should probably focus in this area
  • 7. 7All content copyright © 2019 QuantumBlack, a McKinsey company Advanced hyper-parameter optimization methods
  • 8. 8All content copyright © 2019 QuantumBlack, a McKinsey company Advanced HPO methods are only beneficial in certain situations In advanced optimisation methods, we spend more computational time to figure out what set of parameters 𝜃 we are evaluating the performance metric 𝑓ℳ 𝜃 for. As a result, applying these methods only really makes sense if • The number of hyperparameters is very high (𝜃 is of high dimension); or • It is computationally very expensive to evaluate 𝑓ℳ(𝜃) for a single point 𝜃.
  • 9. 9All content copyright © 2019 QuantumBlack, a McKinsey company There are now a wide range of hyper-parameter optimisation methods HPO Variant Multi- objective Multi-fidelity Search Random / grid search Population- based Bayesian optimisation SMAC3: https://github.com/automl/SMAC3 Hyperopt: http://hyperopt.github.io/hyperopt/ Ax: https://github.com/facebook/Ax Tune: https://ray.readthedocs.io/en/latest/tune.html TPOT: https://github.com/EpistasisLab/tpot Tune: https://ray.readthedocs.io/en/latest/tune.html Hyperband: https://github.com/thuijskens/scikit-hyperband Tune: https://ray.readthedocs.io/en/latest/tune.html Ax: https://github.com/facebook/Ax HpBandSter: https://github.com/automl/HpBandSter Fabolas: https://github.com/automl/RoBO MTBO: https://github.com/automl/RoBO
  • 10. 10All content copyright © 2019 QuantumBlack, a McKinsey company BO-HB is a recent method that combines Bayesian optimisation and multi-fidelity methods HPO Variant Multi- objective Multi-fidelity Search Random / grid search Population- based Bayesian optimisation SMAC3: https://github.com/automl/SMAC3 Hyperopt: http://hyperopt.github.io/hyperopt/ Ax: https://github.com/facebook/Ax Tune: https://ray.readthedocs.io/en/latest/tune.html TPOT: https://github.com/EpistasisLab/tpot Tune: https://ray.readthedocs.io/en/latest/tune.html Hyperband: https://github.com/thuijskens/scikit-hyperband Tune: https://ray.readthedocs.io/en/latest/tune.html Ax: https://github.com/facebook/Ax HpBandSter: https://github.com/automl/HpBandSter Fabolas: https://github.com/automl/RoBO MTBO: https://github.com/automl/RoBO
  • 11. 11All content copyright © 2019 QuantumBlack, a McKinsey company Hyperband
  • 12. 12All content copyright © 2019 QuantumBlack, a McKinsey company Hyperband iteratively runs the successive halving subroutine on model configurations BOHB: Robust and efficient hyperparameter optimization at scale: https://www.automl.org/blog_bohb/
  • 13. 13All content copyright © 2019 QuantumBlack, a McKinsey company Successive halving suffers from the exploitation versus exploration trade-off Given a total computational budget, the user has to decide whether • To try many configurations with a small budget each; or • To try only a few configurations with a larger budget. Hyperband hedges for this trade-off by: • Dividing the total computational budget into several combinations of number of configurations vs. budget. • Call successive halving as a subroutine for each combination.
  • 14. 14All content copyright © 2019 QuantumBlack, a McKinsey company Computational budget Number of configurations Successive halving iteration Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization, Lisha Li, Kevin Jamieson, Giulia DeSalvo, Afshin Rostamizadeh, Ameet Talwalkar, 2018. https://arxiv.org/abs/1603.06560
  • 15. 15All content copyright © 2019 QuantumBlack, a McKinsey company Hyperband’s hedging strategy is very competitive against random search BOHB: Robust and Efficient Hyperparameter Optimization at Scale, Stefan Falkner, Aaron Klein, Frank Hutter, 2018. https://arxiv.org/abs/1807.01774
  • 16. 16All content copyright © 2019 QuantumBlack, a McKinsey company Bayesian optimisation
  • 17. 17All content copyright © 2019 QuantumBlack, a McKinsey company The classic Bayesian optimization algorithm consists of three steps Evaluate performance of learner f with parameters 𝜃 Update current belief of loss surface of the learner f Choose 𝜃 that maximises some utility over the current belief 𝑓 | 𝑦)*+ 𝑦)*+ = 𝑓 𝜃)*+ 𝜃)*+
  • 18. 18All content copyright © 2019 QuantumBlack, a McKinsey company The performance function 𝑓ℳ is modelled to form a current belief of what the loss surface looks like.. Evaluate performance of learner f with parameters 𝜃 Update current belief of loss surface of the learner f Choose 𝜃 that maximises some utility over the current belief 𝑓 | 𝑦)*+ 𝑦)*+ = 𝑓 𝜃)*+ 𝜃)*+
  • 19. 19All content copyright © 2019 QuantumBlack, a McKinsey company ..acquisition functions are used to formalize what constitutes a ”best guess”.. Evaluate performance of learner f with parameters 𝜃 Update current belief of loss surface of the learner f Choose 𝜃 that maximises some utility over the current belief 𝑓 | 𝑦)*+ 𝑦)*+ = 𝑓 𝜃)*+ 𝜃)*+
  • 20. 20All content copyright © 2019 QuantumBlack, a McKinsey company The expected improvement acquisition trades off exploitation of known optimal areas, versus exploration of unexplored areas of the loss surface Exploitation of known optimal space Exploration of unknown space A Tutorial on Bayesian Optimization of Expensive Cost Functions, with Application to Active User Modeling and Hierarchical Reinforcement Learning, Eric Brochu, Vlad M. Cora, Nando de Freitas, 2010. https://arxiv.org/abs/1012.2599
  • 21. 21All content copyright © 2019 QuantumBlack, a McKinsey company .. and finally the performance is computed for suggested parameters Evaluate performance of learner f with parameters 𝜃 Update current belief of loss surface of the learner f Choose 𝜃 that maximises some utility over the current belief 𝑓 | 𝑦)*+ 𝑦)*+ = 𝑓 𝜃)*+ 𝜃)*+
  • 22. 22All content copyright © 2019 QuantumBlack, a McKinsey company Bayesian optimization is competitive against random search for longer training budgets BOHB: Robust and Efficient Hyperparameter Optimization at Scale, Stefan Falkner, Aaron Klein, Frank Hutter, 2018. https://arxiv.org/abs/1807.01774
  • 23. 23All content copyright © 2019 QuantumBlack, a McKinsey company Bayesian optimisation works well for larger budgets, whereas hyperband works better for small to medium budgets BOHB: Robust and Efficient Hyperparameter Optimization at Scale, Stefan Falkner, Aaron Klein, Frank Hutter, 2018. https://arxiv.org/abs/1807.01774
  • 24. 24All content copyright © 2019 QuantumBlack, a McKinsey company BO-HB replaces random search in Hyperband with Bayesian optimisation BOHB: Robust and Efficient Hyperparameter Optimization at Scale, Stefan Falkner, Aaron Klein, Frank Hutter, 2018. https://arxiv.org/abs/1807.01774
  • 25. 25All content copyright © 2019 QuantumBlack, a McKinsey company What HPO method should I use in practice? • If it is possible to define substantially cheaper version of the performance function 𝑓ℳ: Bayesian optimisation-hyperband (BO-HB) • If this is not possible and • if all hyperparameters are real-valued and one can only afford a few dozen function evaluations: Gaussian-process based Bayesian optimisation (SMAC, Spearmint) • if the configuration space is large and/or conditional: Random forest-based Bayesian optimisation (hyperopt)