SlideShare a Scribd company logo
1 of 38
Download to read offline
Managing Machine Learning
David Murgatroyd - VP, Engineering
@dmurga
3
4
5
Problem: I can’t fully specify the behavior I want.
6
Problem: I can’t fully specify the behavior I want.
Solution: Machine Learning
Where does
machine learning fit in
the technology
universe?
Valuable
... a star of the Data Science orchestra.
- John Mount, Win-Vector
Central
... the new algorithms ... at the heart of most of
what computer science does.
- Hal Daumé III, U. Maryland Professor
Last Resort
… for cases when the desired behavior cannot
be effectively expressed in software logic
without dependency on external data.
- D. Sculley et al., Google
7
Where does machine learning fit in developing
technology?
8
Stuff to do Demonstrable ValueStuff to do now
How does machine
learning affect value
demonstration?
Distill business goal into a repeatable,
balanced metric.
Measure on the most representative data you
can get.
Distinguish intrinsic errors from
implementation bugs.
Let your customer override the model when
they absolutely must get some answer.
9
Demonstrable Value
Distill business goal into
a repeatable, balanced
metric.
10
Demonstrable Value
Business goals in our example:
● fewer incorrect candidates sent to
analysts for review
● no increased volume of work for
analysts
● confidence to help analysts prioritize
Example metric: area under an error trade-off
curve based on confidence, constrained to
max volume. Sometimes called an ‘overall
evaluation criteria’ (OEC).
Note that the more skewed the OEC (e.g., if #
of positives varies by day and season) the
more samples are required to be sure of
statistical significance.
Measure on the most
representative data you
can get.
11
Demonstrable Value
Considerations when selecting data:
● online v offline: A/B test in production with
feature flags (one or two variables at a time,
agile-y) vs. stable data set
● implicit v explicit: implicit can correlate
more with value but omits unseen states
● broad v targeted: if explicitly annotating
consider targeting based on diagnostic
value or where systems disagree
Resist the temptation to ‘clean’ data -- you may
kill it. Instead include normalization in your
model.
Distinguish intrinsic
errors from
implementation bugs.
12
Demonstrable Value
Distinction
● Error: incorrect output from a model
despite the model being correctly
implemented.
● Bug: incorrect implementation, doing
something other than what was
intended
Useful to manage expectations about quality
and effort required to improve/fix.
Providing an explanation for output can help
make this distinction.
Bug Error
Let your customer
override the model
when they absolutely
must get some answer.
13
Demonstrable Value
Varieties of overrides:
● Always give this answer.
● Never give this answer.
Can apply for sub-models or overall.
Beware of potential toward ‘whack a mole’.
Feel sad every time they use it.
Where does machine learning fit in developing
technology?
14
Stuff to do Demonstrable ValueStuff to do now
How does machine
learning affect team
organization?
15
Machine Learning Expert
Spectrum of options between:
Integrate machine learning expertise in every
team that needs it.
Separate it in an independent, specialist
team.
Option 1: integrated
teams with cross-team
interest groups
16
Encourages alignment with
business goals.
Challenges machine learning
collaboration, depth and reuse.
Best for small, diverse products.
Option 2: independent
machine learning team
delivering models
17
Encourages machine learning
collaboration, depth and reuse.
Challenges alignment with
business goals.
Best for products with large,
complex model(s).
How does machine
learning affect iteration
structure?
18
Pros for shorter:
● More simple experiments are better
than fewer complex ones
● The value of machine learning leads to
high cost of delay
Pros for longer:
● Innovation takes deep thinking
● More time to control technical debt
creation
Where does machine learning fit in developing
technology?
19
Stuff to do Demonstrable ValueStuff to do now
How does machine
learning affect chunks
of work?
Focus on experiments following the
scientific-method: hypothesis, measurement
and error analysis.
Continuously test for regression versus
expected measurements.
Decouple functional tests from model
variations.
20
Stuff to do now
Focus on experiments
with hypothesis,
measurement and
analysis.
21
Stuff to do now
Continuously test for
regression versus
expected
measurements.
22
Stuff to do now
With machine learning’s dependence on data
changing anything changes everything. This
makes it the “high-interest credit card of
technical debt”.
Determine what’s a significant change,
including looking at aggregate effect across
different data sets.
Decouple functional
tests from model
variations.
23
Stuff to do now
Options:
Black-box style: enforce “can’t be wrong”
(“earmark”) input/output pairs. Might lead to
spurious test failures.
Clear-box style: use a mock implementation
of the model that produces expected answers.
Decouple functional
tests from model
variations.
24
Stuff to do now
Options:
Black-box style: ensure “can’t be wrong”
(“earmark”) input/output pairs. Might lead to
spurious test failures.
Clear-box style: use a mock implementation
of the model that produces expected answers.
Decouple functional
tests from model
variations.
25
Stuff to do now
Options:
Black-box style: ensure “can’t be wrong”
(“earmark”) input/output pairs. Might lead to
spurious test failures.
Clear-box style: use a mock implementation
of the model that produces expected answers.
42
Where does machine learning fit in developing
technology?
26
Stuff to do Demonstrable ValueStuff to do now
How does machine
learning affect
prioritization?
27
Stuff to do
Do we need more training data?
Do we need a richer representation of our
data?
Do we need a combination of models?
How much could improving a sub-component
of the model help?
What development milestones should we
target?
Do we need more
training data?
28
Stuff to do
The learning curve implies adding training
data should bring down the test error closer
to the desired level.
Do we need a richer
representation of our
data?
29
Stuff to do
The learning curve implies adding data won’t
help but a richer data representation may.
Could be more features identified by
someone with domain expertise analyzing
errors. Though remember more features
often means less speed.
Could require a new model if the domain
information identified is not representable in
the existing one.
Do we need a
combination of models?
30
Stuff to do
The learning curve implies the model is
overfitting the training set.
Consider training multiple models on random
subsets of the data and combine them at
runtime to decrease the variance while
retaining a low bias. Presuming you can spend
the compute.
How much could
improving a
sub-component of the
model help?
31
Stuff to do
Build an ‘oracle’ for the sub-component --
something that takes perfect output from
data.
Annotate to get that perfect output on some
test data to feed the oracle.
Measure the overall system with the oracle
turned on.
What development
milestones should we
target?
32
Stuff to do
Make it…
● Glued-together with some rules
(Prototype)
● Function (Alpha)
● Measurable & inspectable (early Beta)
● Accurate, not slow, nice demo,
documented & configurable (late Beta)
● Simple & fast (GA)
● Handle new kinds of input (post-GA)
Questions?
33
Stuff to do Demonstrable ValueStuff to do now
Suggested questions:
Say more about integrating domain expertise?
Say more about online vs. offline testing?
How to manage acquiring data?
How to recruit machine learning folks?
What bad habits can ML enable?
Where can I try your stuff? api.rosette.com
You hiring? Yes - basistech.com/careers/
@dmurga
Appendix
@dmurga
Manage Technical Debt
35
Data is even less visible than code -- blurs
boundaries.
Recruiting machine learning experts
36
who
◦ expertise in sequence models > in domain
◦ depth in specific model > breadth over many
where to find them
◦ local network: meet-ups, LinkedIn
◦ academic conferences
◦ communities (e.g., Kaggle, users of ML tools)
how to attract them
◦ explain purpose & uniqueness of the problem
Online vs. offline evaluation
37
Online (e.g., A/B)
● Individual decisions need to not be mission critical
● Enough use to get sufficient statistics in short time
● Helps motivate aligning production and development environments
● If the model is updated online, validate it against offline data periodically to
watch out for drift
● Usually focused on extrinsic or distant measures
Offline
● Always have some of this to for long-term protection against regression
● May be required for intrinsic measurement
38
Epistemology Exact
sciences
Experimental
sciences
Engineering Art
Example ... Theoretical C.S. Physics Software Management
Deals with ... Theorems Theories Artifacts People
Truth is ... Forever Temporary “It works” In the eye of
the beholder
Parts of
machine
learning fit all
four...
Learning theory Model &
measure
Systems Users
This is great, as long as we don’t confuse one kind of work for another.
(This table is an expansion of one in Bottou’s ICML 2015 talk.)

More Related Content

What's hot

Barga Data Science lecture 1
Barga Data Science lecture 1Barga Data Science lecture 1
Barga Data Science lecture 1Roger Barga
 
Barga DIDC'14 Invited Talk
Barga DIDC'14 Invited TalkBarga DIDC'14 Invited Talk
Barga DIDC'14 Invited TalkRoger Barga
 
Barga Data Science lecture 4
Barga Data Science lecture 4Barga Data Science lecture 4
Barga Data Science lecture 4Roger Barga
 
Introduction to machine learning
Introduction to machine learningIntroduction to machine learning
Introduction to machine learningKoundinya Desiraju
 
Top 10 Data Science Practitioner Pitfalls
Top 10 Data Science Practitioner PitfallsTop 10 Data Science Practitioner Pitfalls
Top 10 Data Science Practitioner PitfallsSri Ambati
 
End-to-End Machine Learning Project
End-to-End Machine Learning ProjectEnd-to-End Machine Learning Project
End-to-End Machine Learning ProjectEng Teong Cheah
 
H2O World - Top 10 Data Science Pitfalls - Mark Landry
H2O World - Top 10 Data Science Pitfalls - Mark LandryH2O World - Top 10 Data Science Pitfalls - Mark Landry
H2O World - Top 10 Data Science Pitfalls - Mark LandrySri Ambati
 
Barga Data Science lecture 5
Barga Data Science lecture 5Barga Data Science lecture 5
Barga Data Science lecture 5Roger Barga
 
Machine learning the next revolution or just another hype
Machine learning   the next revolution or just another hypeMachine learning   the next revolution or just another hype
Machine learning the next revolution or just another hypeJorge Ferrer
 
Introduction to Data Science
Introduction to Data ScienceIntroduction to Data Science
Introduction to Data ScienceNiko Vuokko
 
Barga Data Science lecture 10
Barga Data Science lecture 10Barga Data Science lecture 10
Barga Data Science lecture 10Roger Barga
 
Machine Learning: Understanding the Invisible Force Changing Our World
Machine Learning: Understanding the Invisible Force Changing Our WorldMachine Learning: Understanding the Invisible Force Changing Our World
Machine Learning: Understanding the Invisible Force Changing Our WorldKen Tabor
 
Barga Data Science lecture 7
Barga Data Science lecture 7Barga Data Science lecture 7
Barga Data Science lecture 7Roger Barga
 
Fairly Measuring Fairness In Machine Learning
Fairly Measuring Fairness In Machine LearningFairly Measuring Fairness In Machine Learning
Fairly Measuring Fairness In Machine LearningHJ van Veen
 
Machine learning basics
Machine learning basics Machine learning basics
Machine learning basics Akanksha Bali
 
Machine learning_ Replicating Human Brain
Machine learning_ Replicating Human BrainMachine learning_ Replicating Human Brain
Machine learning_ Replicating Human BrainNishant Jain
 
10 Lessons Learned from Building Machine Learning Systems
10 Lessons Learned from Building Machine Learning Systems10 Lessons Learned from Building Machine Learning Systems
10 Lessons Learned from Building Machine Learning SystemsXavier Amatriain
 
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...Aseda Owusua Addai-Deseh
 

What's hot (20)

Barga Data Science lecture 1
Barga Data Science lecture 1Barga Data Science lecture 1
Barga Data Science lecture 1
 
Barga DIDC'14 Invited Talk
Barga DIDC'14 Invited TalkBarga DIDC'14 Invited Talk
Barga DIDC'14 Invited Talk
 
Barga Data Science lecture 4
Barga Data Science lecture 4Barga Data Science lecture 4
Barga Data Science lecture 4
 
Agile Deep Learning
Agile Deep LearningAgile Deep Learning
Agile Deep Learning
 
Introduction to machine learning
Introduction to machine learningIntroduction to machine learning
Introduction to machine learning
 
Top 10 Data Science Practitioner Pitfalls
Top 10 Data Science Practitioner PitfallsTop 10 Data Science Practitioner Pitfalls
Top 10 Data Science Practitioner Pitfalls
 
End-to-End Machine Learning Project
End-to-End Machine Learning ProjectEnd-to-End Machine Learning Project
End-to-End Machine Learning Project
 
H2O World - Top 10 Data Science Pitfalls - Mark Landry
H2O World - Top 10 Data Science Pitfalls - Mark LandryH2O World - Top 10 Data Science Pitfalls - Mark Landry
H2O World - Top 10 Data Science Pitfalls - Mark Landry
 
Barga Data Science lecture 5
Barga Data Science lecture 5Barga Data Science lecture 5
Barga Data Science lecture 5
 
Machine Learning Algorithms (Part 1)
Machine Learning Algorithms (Part 1)Machine Learning Algorithms (Part 1)
Machine Learning Algorithms (Part 1)
 
Machine learning the next revolution or just another hype
Machine learning   the next revolution or just another hypeMachine learning   the next revolution or just another hype
Machine learning the next revolution or just another hype
 
Introduction to Data Science
Introduction to Data ScienceIntroduction to Data Science
Introduction to Data Science
 
Barga Data Science lecture 10
Barga Data Science lecture 10Barga Data Science lecture 10
Barga Data Science lecture 10
 
Machine Learning: Understanding the Invisible Force Changing Our World
Machine Learning: Understanding the Invisible Force Changing Our WorldMachine Learning: Understanding the Invisible Force Changing Our World
Machine Learning: Understanding the Invisible Force Changing Our World
 
Barga Data Science lecture 7
Barga Data Science lecture 7Barga Data Science lecture 7
Barga Data Science lecture 7
 
Fairly Measuring Fairness In Machine Learning
Fairly Measuring Fairness In Machine LearningFairly Measuring Fairness In Machine Learning
Fairly Measuring Fairness In Machine Learning
 
Machine learning basics
Machine learning basics Machine learning basics
Machine learning basics
 
Machine learning_ Replicating Human Brain
Machine learning_ Replicating Human BrainMachine learning_ Replicating Human Brain
Machine learning_ Replicating Human Brain
 
10 Lessons Learned from Building Machine Learning Systems
10 Lessons Learned from Building Machine Learning Systems10 Lessons Learned from Building Machine Learning Systems
10 Lessons Learned from Building Machine Learning Systems
 
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
Day 2 (Lecture 5): A Practitioner's Perspective on Building Machine Product i...
 

Viewers also liked

キーボードアプリとSketchのススメ
キーボードアプリとSketchのススメキーボードアプリとSketchのススメ
キーボードアプリとSketchのススメYuichi Yoshida
 
Lessons learned
Lessons learnedLessons learned
Lessons learnedhexgnu
 
BAHSICアルゴリズムによる非線形データからの特徴選択
BAHSICアルゴリズムによる非線形データからの特徴選択BAHSICアルゴリズムによる非線形データからの特徴選択
BAHSICアルゴリズムによる非線形データからの特徴選択Mika Yoshimura
 
Linguistic Considerations of Identity Resolution (2008)
Linguistic Considerations of Identity Resolution (2008)Linguistic Considerations of Identity Resolution (2008)
Linguistic Considerations of Identity Resolution (2008)David Murgatroyd
 
Machine Learning : The high interest credit card of technical debt
Machine Learning : The high interest credit card of technical debt Machine Learning : The high interest credit card of technical debt
Machine Learning : The high interest credit card of technical debt Yuichi Yoshida
 
C/C++プログラマのための開発ツール
C/C++プログラマのための開発ツールC/C++プログラマのための開発ツール
C/C++プログラマのための開発ツールMITSUNARI Shigeo
 
汎用性と高速性を目指したペアリング暗号ライブラリ mcl
汎用性と高速性を目指したペアリング暗号ライブラリ mcl汎用性と高速性を目指したペアリング暗号ライブラリ mcl
汎用性と高速性を目指したペアリング暗号ライブラリ mclMITSUNARI Shigeo
 
「機械学習:技術的負債の高利子クレジットカード」のまとめ
「機械学習:技術的負債の高利子クレジットカード」のまとめ「機械学習:技術的負債の高利子クレジットカード」のまとめ
「機械学習:技術的負債の高利子クレジットカード」のまとめRecruit Technologies
 
Let's Make the PAIN Visible!
Let's Make the PAIN Visible!Let's Make the PAIN Visible!
Let's Make the PAIN Visible!Arty Starr
 
A Programmer's Guide to Humans
A Programmer's Guide to HumansA Programmer's Guide to Humans
A Programmer's Guide to HumansArty Starr
 
GoogleのSHA-1のはなし
GoogleのSHA-1のはなしGoogleのSHA-1のはなし
GoogleのSHA-1のはなしMITSUNARI Shigeo
 

Viewers also liked (11)

キーボードアプリとSketchのススメ
キーボードアプリとSketchのススメキーボードアプリとSketchのススメ
キーボードアプリとSketchのススメ
 
Lessons learned
Lessons learnedLessons learned
Lessons learned
 
BAHSICアルゴリズムによる非線形データからの特徴選択
BAHSICアルゴリズムによる非線形データからの特徴選択BAHSICアルゴリズムによる非線形データからの特徴選択
BAHSICアルゴリズムによる非線形データからの特徴選択
 
Linguistic Considerations of Identity Resolution (2008)
Linguistic Considerations of Identity Resolution (2008)Linguistic Considerations of Identity Resolution (2008)
Linguistic Considerations of Identity Resolution (2008)
 
Machine Learning : The high interest credit card of technical debt
Machine Learning : The high interest credit card of technical debt Machine Learning : The high interest credit card of technical debt
Machine Learning : The high interest credit card of technical debt
 
C/C++プログラマのための開発ツール
C/C++プログラマのための開発ツールC/C++プログラマのための開発ツール
C/C++プログラマのための開発ツール
 
汎用性と高速性を目指したペアリング暗号ライブラリ mcl
汎用性と高速性を目指したペアリング暗号ライブラリ mcl汎用性と高速性を目指したペアリング暗号ライブラリ mcl
汎用性と高速性を目指したペアリング暗号ライブラリ mcl
 
「機械学習:技術的負債の高利子クレジットカード」のまとめ
「機械学習:技術的負債の高利子クレジットカード」のまとめ「機械学習:技術的負債の高利子クレジットカード」のまとめ
「機械学習:技術的負債の高利子クレジットカード」のまとめ
 
Let's Make the PAIN Visible!
Let's Make the PAIN Visible!Let's Make the PAIN Visible!
Let's Make the PAIN Visible!
 
A Programmer's Guide to Humans
A Programmer's Guide to HumansA Programmer's Guide to Humans
A Programmer's Guide to Humans
 
GoogleのSHA-1のはなし
GoogleのSHA-1のはなしGoogleのSHA-1のはなし
GoogleのSHA-1のはなし
 

Similar to Managing machine learning

Course 2 Machine Learning Data LifeCycle in Production - Week 1
Course 2   Machine Learning Data LifeCycle in Production - Week 1Course 2   Machine Learning Data LifeCycle in Production - Week 1
Course 2 Machine Learning Data LifeCycle in Production - Week 1Ajay Taneja
 
2024-02-24_Session 1 - PMLE_UPDATED.pptx
2024-02-24_Session 1 - PMLE_UPDATED.pptx2024-02-24_Session 1 - PMLE_UPDATED.pptx
2024-02-24_Session 1 - PMLE_UPDATED.pptxgdgsurrey
 
Data Science for Business Managers - An intro to ROI for predictive analytics
Data Science for Business Managers - An intro to ROI for predictive analyticsData Science for Business Managers - An intro to ROI for predictive analytics
Data Science for Business Managers - An intro to ROI for predictive analyticsAkin Osman Kazakci
 
The 4 Machine Learning Models Imperative for Business Transformation
The 4 Machine Learning Models Imperative for Business TransformationThe 4 Machine Learning Models Imperative for Business Transformation
The 4 Machine Learning Models Imperative for Business TransformationRocketSource
 
Machine Learning vs Decision Optimization comparison
Machine Learning vs Decision Optimization comparisonMachine Learning vs Decision Optimization comparison
Machine Learning vs Decision Optimization comparisonAlain Chabrier
 
(In)convenient truths about applied machine learning
(In)convenient truths about applied machine learning(In)convenient truths about applied machine learning
(In)convenient truths about applied machine learningMax Pagels
 
Afternoons with Azure - Azure Machine Learning
Afternoons with Azure - Azure Machine Learning Afternoons with Azure - Azure Machine Learning
Afternoons with Azure - Azure Machine Learning CCG
 
Lead AI incubations as a Product manager
Lead AI incubations as a Product manager Lead AI incubations as a Product manager
Lead AI incubations as a Product manager Debapriya Basu
 
AI-900 - Fundamental Principles of ML.pptx
AI-900 - Fundamental Principles of ML.pptxAI-900 - Fundamental Principles of ML.pptx
AI-900 - Fundamental Principles of ML.pptxkprasad8
 
Putting data science in your business a first utility feedback
Putting data science in your business a first utility feedbackPutting data science in your business a first utility feedback
Putting data science in your business a first utility feedbackPeculium Crypto
 
How Will Your ML Project Fail
How Will Your ML Project FailHow Will Your ML Project Fail
How Will Your ML Project FailElena Samuylova
 
AI TESTING: ENSURING A GOOD DATA SPLIT BETWEEN DATA SETS (TRAINING AND TEST) ...
AI TESTING: ENSURING A GOOD DATA SPLIT BETWEEN DATA SETS (TRAINING AND TEST) ...AI TESTING: ENSURING A GOOD DATA SPLIT BETWEEN DATA SETS (TRAINING AND TEST) ...
AI TESTING: ENSURING A GOOD DATA SPLIT BETWEEN DATA SETS (TRAINING AND TEST) ...ijsc
 
Managing uncertainty in ai performance target setting
Managing uncertainty in ai performance target settingManaging uncertainty in ai performance target setting
Managing uncertainty in ai performance target settingNoelle Ibrahim
 
Managing uncertainty in ai performance target setting
Managing uncertainty in ai performance target settingManaging uncertainty in ai performance target setting
Managing uncertainty in ai performance target settingNoelle Ibrahim
 
Choose the Right Problems to Solve with ML by Spotify PM
Choose the Right Problems to Solve with ML by Spotify PMChoose the Right Problems to Solve with ML by Spotify PM
Choose the Right Problems to Solve with ML by Spotify PMProduct School
 
Continuous Intelligence Workshop
Continuous Intelligence WorkshopContinuous Intelligence Workshop
Continuous Intelligence WorkshopDavid Tan
 
Machine intelligence data science methodology 060420
Machine intelligence data science methodology 060420Machine intelligence data science methodology 060420
Machine intelligence data science methodology 060420Jeremy Lehman
 
Machine learning: A Walk Through School Exams
Machine learning: A Walk Through School ExamsMachine learning: A Walk Through School Exams
Machine learning: A Walk Through School ExamsRamsha Ijaz
 

Similar to Managing machine learning (20)

Course 2 Machine Learning Data LifeCycle in Production - Week 1
Course 2   Machine Learning Data LifeCycle in Production - Week 1Course 2   Machine Learning Data LifeCycle in Production - Week 1
Course 2 Machine Learning Data LifeCycle in Production - Week 1
 
2024-02-24_Session 1 - PMLE_UPDATED.pptx
2024-02-24_Session 1 - PMLE_UPDATED.pptx2024-02-24_Session 1 - PMLE_UPDATED.pptx
2024-02-24_Session 1 - PMLE_UPDATED.pptx
 
Data Science for Business Managers - An intro to ROI for predictive analytics
Data Science for Business Managers - An intro to ROI for predictive analyticsData Science for Business Managers - An intro to ROI for predictive analytics
Data Science for Business Managers - An intro to ROI for predictive analytics
 
Demystifying ML/AI
Demystifying ML/AIDemystifying ML/AI
Demystifying ML/AI
 
The 4 Machine Learning Models Imperative for Business Transformation
The 4 Machine Learning Models Imperative for Business TransformationThe 4 Machine Learning Models Imperative for Business Transformation
The 4 Machine Learning Models Imperative for Business Transformation
 
Machine Learning vs Decision Optimization comparison
Machine Learning vs Decision Optimization comparisonMachine Learning vs Decision Optimization comparison
Machine Learning vs Decision Optimization comparison
 
(In)convenient truths about applied machine learning
(In)convenient truths about applied machine learning(In)convenient truths about applied machine learning
(In)convenient truths about applied machine learning
 
Afternoons with Azure - Azure Machine Learning
Afternoons with Azure - Azure Machine Learning Afternoons with Azure - Azure Machine Learning
Afternoons with Azure - Azure Machine Learning
 
Lead AI incubations as a Product manager
Lead AI incubations as a Product manager Lead AI incubations as a Product manager
Lead AI incubations as a Product manager
 
AI-900 - Fundamental Principles of ML.pptx
AI-900 - Fundamental Principles of ML.pptxAI-900 - Fundamental Principles of ML.pptx
AI-900 - Fundamental Principles of ML.pptx
 
Putting data science in your business a first utility feedback
Putting data science in your business a first utility feedbackPutting data science in your business a first utility feedback
Putting data science in your business a first utility feedback
 
How Will Your ML Project Fail
How Will Your ML Project FailHow Will Your ML Project Fail
How Will Your ML Project Fail
 
AI TESTING: ENSURING A GOOD DATA SPLIT BETWEEN DATA SETS (TRAINING AND TEST) ...
AI TESTING: ENSURING A GOOD DATA SPLIT BETWEEN DATA SETS (TRAINING AND TEST) ...AI TESTING: ENSURING A GOOD DATA SPLIT BETWEEN DATA SETS (TRAINING AND TEST) ...
AI TESTING: ENSURING A GOOD DATA SPLIT BETWEEN DATA SETS (TRAINING AND TEST) ...
 
Quality myths
Quality mythsQuality myths
Quality myths
 
Managing uncertainty in ai performance target setting
Managing uncertainty in ai performance target settingManaging uncertainty in ai performance target setting
Managing uncertainty in ai performance target setting
 
Managing uncertainty in ai performance target setting
Managing uncertainty in ai performance target settingManaging uncertainty in ai performance target setting
Managing uncertainty in ai performance target setting
 
Choose the Right Problems to Solve with ML by Spotify PM
Choose the Right Problems to Solve with ML by Spotify PMChoose the Right Problems to Solve with ML by Spotify PM
Choose the Right Problems to Solve with ML by Spotify PM
 
Continuous Intelligence Workshop
Continuous Intelligence WorkshopContinuous Intelligence Workshop
Continuous Intelligence Workshop
 
Machine intelligence data science methodology 060420
Machine intelligence data science methodology 060420Machine intelligence data science methodology 060420
Machine intelligence data science methodology 060420
 
Machine learning: A Walk Through School Exams
Machine learning: A Walk Through School ExamsMachine learning: A Walk Through School Exams
Machine learning: A Walk Through School Exams
 

More from David Murgatroyd

Mission-Driven Machine Learning
Mission-Driven Machine LearningMission-Driven Machine Learning
Mission-Driven Machine LearningDavid Murgatroyd
 
Leveraging AI the Right Way (for Product Managers)
Leveraging AI the Right Way (for Product Managers)Leveraging AI the Right Way (for Product Managers)
Leveraging AI the Right Way (for Product Managers)David Murgatroyd
 
Managing Your Machine Learning Portfolio
Managing Your Machine Learning PortfolioManaging Your Machine Learning Portfolio
Managing Your Machine Learning PortfolioDavid Murgatroyd
 
How to train your product owner
How to train your product ownerHow to train your product owner
How to train your product ownerDavid Murgatroyd
 
Technology & Faith: from Coding to Culture
Technology & Faith: from Coding to CultureTechnology & Faith: from Coding to Culture
Technology & Faith: from Coding to CultureDavid Murgatroyd
 
Choosing a Job for the Right Reasons
Choosing a Job for the Right ReasonsChoosing a Job for the Right Reasons
Choosing a Job for the Right ReasonsDavid Murgatroyd
 
System combination for HLT
System combination for HLTSystem combination for HLT
System combination for HLTDavid Murgatroyd
 
Simple fuzzy name matching in solr
Simple fuzzy name matching in solrSimple fuzzy name matching in solr
Simple fuzzy name matching in solrDavid Murgatroyd
 
Moving beyond-entity-extraction-to-entity-resolution-david-murgatroyd-human-l...
Moving beyond-entity-extraction-to-entity-resolution-david-murgatroyd-human-l...Moving beyond-entity-extraction-to-entity-resolution-david-murgatroyd-human-l...
Moving beyond-entity-extraction-to-entity-resolution-david-murgatroyd-human-l...David Murgatroyd
 
From Research to Reality: Advances in HLT 2013
From Research to Reality: Advances in HLT 2013From Research to Reality: Advances in HLT 2013
From Research to Reality: Advances in HLT 2013David Murgatroyd
 

More from David Murgatroyd (12)

Mission-Driven Machine Learning
Mission-Driven Machine LearningMission-Driven Machine Learning
Mission-Driven Machine Learning
 
Leveraging AI the Right Way (for Product Managers)
Leveraging AI the Right Way (for Product Managers)Leveraging AI the Right Way (for Product Managers)
Leveraging AI the Right Way (for Product Managers)
 
Managing Your Machine Learning Portfolio
Managing Your Machine Learning PortfolioManaging Your Machine Learning Portfolio
Managing Your Machine Learning Portfolio
 
How to train your product owner
How to train your product ownerHow to train your product owner
How to train your product owner
 
Technology & Faith: from Coding to Culture
Technology & Faith: from Coding to CultureTechnology & Faith: from Coding to Culture
Technology & Faith: from Coding to Culture
 
Choosing a Job for the Right Reasons
Choosing a Job for the Right ReasonsChoosing a Job for the Right Reasons
Choosing a Job for the Right Reasons
 
NLP in the Real World
NLP in the Real WorldNLP in the Real World
NLP in the Real World
 
System combination for HLT
System combination for HLTSystem combination for HLT
System combination for HLT
 
HltCon overview
HltCon overviewHltCon overview
HltCon overview
 
Simple fuzzy name matching in solr
Simple fuzzy name matching in solrSimple fuzzy name matching in solr
Simple fuzzy name matching in solr
 
Moving beyond-entity-extraction-to-entity-resolution-david-murgatroyd-human-l...
Moving beyond-entity-extraction-to-entity-resolution-david-murgatroyd-human-l...Moving beyond-entity-extraction-to-entity-resolution-david-murgatroyd-human-l...
Moving beyond-entity-extraction-to-entity-resolution-david-murgatroyd-human-l...
 
From Research to Reality: Advances in HLT 2013
From Research to Reality: Advances in HLT 2013From Research to Reality: Advances in HLT 2013
From Research to Reality: Advances in HLT 2013
 

Recently uploaded

Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...DianaGray10
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationRadu Cotescu
 
presentation ICT roal in 21st century education
presentation ICT roal in 21st century educationpresentation ICT roal in 21st century education
presentation ICT roal in 21st century educationjfdjdjcjdnsjd
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Miguel Araújo
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processorsdebabhi2
 
AWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAndrey Devyatkin
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonAnna Loughnan Colquhoun
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...apidays
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProduct Anonymous
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024The Digital Insurer
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfsudhanshuwaghmare1
 
Developing An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilDeveloping An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilV3cube
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationSafe Software
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...Martijn de Jong
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoffsammart93
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherRemote DBA Services
 

Recently uploaded (20)

Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
Connector Corner: Accelerate revenue generation using UiPath API-centric busi...
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organization
 
presentation ICT roal in 21st century education
presentation ICT roal in 21st century educationpresentation ICT roal in 21st century education
presentation ICT roal in 21st century education
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processors
 
AWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of Terraform
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt Robison
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemkeProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
ProductAnonymous-April2024-WinProductDiscovery-MelissaKlemke
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 
Developing An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of BrazilDeveloping An App To Navigate The Roads of Brazil
Developing An App To Navigate The Roads of Brazil
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a Fresher
 

Managing machine learning

  • 1. Managing Machine Learning David Murgatroyd - VP, Engineering @dmurga
  • 2.
  • 3. 3
  • 4. 4
  • 5. 5 Problem: I can’t fully specify the behavior I want.
  • 6. 6 Problem: I can’t fully specify the behavior I want. Solution: Machine Learning
  • 7. Where does machine learning fit in the technology universe? Valuable ... a star of the Data Science orchestra. - John Mount, Win-Vector Central ... the new algorithms ... at the heart of most of what computer science does. - Hal Daumé III, U. Maryland Professor Last Resort … for cases when the desired behavior cannot be effectively expressed in software logic without dependency on external data. - D. Sculley et al., Google 7
  • 8. Where does machine learning fit in developing technology? 8 Stuff to do Demonstrable ValueStuff to do now
  • 9. How does machine learning affect value demonstration? Distill business goal into a repeatable, balanced metric. Measure on the most representative data you can get. Distinguish intrinsic errors from implementation bugs. Let your customer override the model when they absolutely must get some answer. 9 Demonstrable Value
  • 10. Distill business goal into a repeatable, balanced metric. 10 Demonstrable Value Business goals in our example: ● fewer incorrect candidates sent to analysts for review ● no increased volume of work for analysts ● confidence to help analysts prioritize Example metric: area under an error trade-off curve based on confidence, constrained to max volume. Sometimes called an ‘overall evaluation criteria’ (OEC). Note that the more skewed the OEC (e.g., if # of positives varies by day and season) the more samples are required to be sure of statistical significance.
  • 11. Measure on the most representative data you can get. 11 Demonstrable Value Considerations when selecting data: ● online v offline: A/B test in production with feature flags (one or two variables at a time, agile-y) vs. stable data set ● implicit v explicit: implicit can correlate more with value but omits unseen states ● broad v targeted: if explicitly annotating consider targeting based on diagnostic value or where systems disagree Resist the temptation to ‘clean’ data -- you may kill it. Instead include normalization in your model.
  • 12. Distinguish intrinsic errors from implementation bugs. 12 Demonstrable Value Distinction ● Error: incorrect output from a model despite the model being correctly implemented. ● Bug: incorrect implementation, doing something other than what was intended Useful to manage expectations about quality and effort required to improve/fix. Providing an explanation for output can help make this distinction. Bug Error
  • 13. Let your customer override the model when they absolutely must get some answer. 13 Demonstrable Value Varieties of overrides: ● Always give this answer. ● Never give this answer. Can apply for sub-models or overall. Beware of potential toward ‘whack a mole’. Feel sad every time they use it.
  • 14. Where does machine learning fit in developing technology? 14 Stuff to do Demonstrable ValueStuff to do now
  • 15. How does machine learning affect team organization? 15 Machine Learning Expert Spectrum of options between: Integrate machine learning expertise in every team that needs it. Separate it in an independent, specialist team.
  • 16. Option 1: integrated teams with cross-team interest groups 16 Encourages alignment with business goals. Challenges machine learning collaboration, depth and reuse. Best for small, diverse products.
  • 17. Option 2: independent machine learning team delivering models 17 Encourages machine learning collaboration, depth and reuse. Challenges alignment with business goals. Best for products with large, complex model(s).
  • 18. How does machine learning affect iteration structure? 18 Pros for shorter: ● More simple experiments are better than fewer complex ones ● The value of machine learning leads to high cost of delay Pros for longer: ● Innovation takes deep thinking ● More time to control technical debt creation
  • 19. Where does machine learning fit in developing technology? 19 Stuff to do Demonstrable ValueStuff to do now
  • 20. How does machine learning affect chunks of work? Focus on experiments following the scientific-method: hypothesis, measurement and error analysis. Continuously test for regression versus expected measurements. Decouple functional tests from model variations. 20 Stuff to do now
  • 21. Focus on experiments with hypothesis, measurement and analysis. 21 Stuff to do now
  • 22. Continuously test for regression versus expected measurements. 22 Stuff to do now With machine learning’s dependence on data changing anything changes everything. This makes it the “high-interest credit card of technical debt”. Determine what’s a significant change, including looking at aggregate effect across different data sets.
  • 23. Decouple functional tests from model variations. 23 Stuff to do now Options: Black-box style: enforce “can’t be wrong” (“earmark”) input/output pairs. Might lead to spurious test failures. Clear-box style: use a mock implementation of the model that produces expected answers.
  • 24. Decouple functional tests from model variations. 24 Stuff to do now Options: Black-box style: ensure “can’t be wrong” (“earmark”) input/output pairs. Might lead to spurious test failures. Clear-box style: use a mock implementation of the model that produces expected answers.
  • 25. Decouple functional tests from model variations. 25 Stuff to do now Options: Black-box style: ensure “can’t be wrong” (“earmark”) input/output pairs. Might lead to spurious test failures. Clear-box style: use a mock implementation of the model that produces expected answers. 42
  • 26. Where does machine learning fit in developing technology? 26 Stuff to do Demonstrable ValueStuff to do now
  • 27. How does machine learning affect prioritization? 27 Stuff to do Do we need more training data? Do we need a richer representation of our data? Do we need a combination of models? How much could improving a sub-component of the model help? What development milestones should we target?
  • 28. Do we need more training data? 28 Stuff to do The learning curve implies adding training data should bring down the test error closer to the desired level.
  • 29. Do we need a richer representation of our data? 29 Stuff to do The learning curve implies adding data won’t help but a richer data representation may. Could be more features identified by someone with domain expertise analyzing errors. Though remember more features often means less speed. Could require a new model if the domain information identified is not representable in the existing one.
  • 30. Do we need a combination of models? 30 Stuff to do The learning curve implies the model is overfitting the training set. Consider training multiple models on random subsets of the data and combine them at runtime to decrease the variance while retaining a low bias. Presuming you can spend the compute.
  • 31. How much could improving a sub-component of the model help? 31 Stuff to do Build an ‘oracle’ for the sub-component -- something that takes perfect output from data. Annotate to get that perfect output on some test data to feed the oracle. Measure the overall system with the oracle turned on.
  • 32. What development milestones should we target? 32 Stuff to do Make it… ● Glued-together with some rules (Prototype) ● Function (Alpha) ● Measurable & inspectable (early Beta) ● Accurate, not slow, nice demo, documented & configurable (late Beta) ● Simple & fast (GA) ● Handle new kinds of input (post-GA)
  • 33. Questions? 33 Stuff to do Demonstrable ValueStuff to do now Suggested questions: Say more about integrating domain expertise? Say more about online vs. offline testing? How to manage acquiring data? How to recruit machine learning folks? What bad habits can ML enable? Where can I try your stuff? api.rosette.com You hiring? Yes - basistech.com/careers/ @dmurga
  • 35. Manage Technical Debt 35 Data is even less visible than code -- blurs boundaries.
  • 36. Recruiting machine learning experts 36 who ◦ expertise in sequence models > in domain ◦ depth in specific model > breadth over many where to find them ◦ local network: meet-ups, LinkedIn ◦ academic conferences ◦ communities (e.g., Kaggle, users of ML tools) how to attract them ◦ explain purpose & uniqueness of the problem
  • 37. Online vs. offline evaluation 37 Online (e.g., A/B) ● Individual decisions need to not be mission critical ● Enough use to get sufficient statistics in short time ● Helps motivate aligning production and development environments ● If the model is updated online, validate it against offline data periodically to watch out for drift ● Usually focused on extrinsic or distant measures Offline ● Always have some of this to for long-term protection against regression ● May be required for intrinsic measurement
  • 38. 38 Epistemology Exact sciences Experimental sciences Engineering Art Example ... Theoretical C.S. Physics Software Management Deals with ... Theorems Theories Artifacts People Truth is ... Forever Temporary “It works” In the eye of the beholder Parts of machine learning fit all four... Learning theory Model & measure Systems Users This is great, as long as we don’t confuse one kind of work for another. (This table is an expansion of one in Bottou’s ICML 2015 talk.)

Editor's Notes

  1. Balance: Consistency v correctness Extrinsic v intrinsic Interpretability v correctness Precision v recall (volume) Exploitation v exploration Data: Historic Diagnostic Online v offline
  2. Balance: Consistency v correctness Extrinsic v intrinsic Interpretability v correctness Precision v recall (volume) Exploitation v exploration Data: Historic Diagnostic Online v offline
  3. Balance: Consistency v correctness Extrinsic v intrinsic Interpretability v correctness Precision v recall (volume) Exploitation v exploration Data: Historic Diagnostic Online v offline
  4. Balance: Consistency v correctness Extrinsic v intrinsic Interpretability v correctness Precision v recall (volume) Exploitation v exploration Data: Historic Diagnostic Online v offline
  5. Balance: Consistency v correctness Extrinsic v intrinsic Interpretability v correctness Precision v recall (volume) Exploitation v exploration Data: Historic Diagnostic Online v offline
  6. For online A/B tests, choose control. Oracle Experiments both for data collection and speed (esp of adding caches)
  7. “Changing Anything Changes Everything”