SlideShare une entreprise Scribd logo
1  sur  26
the flight MH370
& Bayes theory
Masilamani Ramasamy
- a scientific management expert
(misilamani@yahoo.com)
Masilamani Ramasamy
Consultant Trainer
MH370 tracking
Bayes' Theorem
• also called, Bayes' law or Bayes' rule
• based on probability theory,
mathematics and Statistics
• named after Thomas Bayes (/ˈbeɪz/
1701–1761)
• Bayes suggested the theorem to
update beliefs
• Significantly edited and updated by
Richard Price
• Bayes' theorem “is to the theory of
probability what Pythagoras's
theorem is to geometry” -Harold
Jeffreys
probability theory
rooted in,
• the science of
Statistics
• the science of
Mathematics
• the probability of an
event occurring
• the analysis of
random phenomena
statistical theory
gives,
–ways of comparing
statistical
procedures
–a best possible
procedure for
statistical problems
–guidance on the
choice between
alternative
procedures
Passenger Identities Number
China/Taiwan 154
Malaysia 38
India 5
Indonesia 7
Australia 6
France 4
USA 3
New Zealand 2
Canada 2
Russia 1
Italy 1
Netherlands 1
Austria 1
statistical theory
covers
– Statistical decision
methods and problems
– Statistical inferences or
conclusions
– actions and deductions
To satisfy the basic principles
stated for these different
approaches.
STATISTICS(FACTS)
Aircraft: Malaysian Air Lines
Aircraft Age: 11years 10
months
Aircraft Type: Boeing 777 ER
Flight No: MH370(shared with
China Southern Airlines)
Passengers travelled: 227
Crew Travelled: 12
Nationalities: 14
Take Off Time: 12.41 am
From: Kuala Lumpur, Malaysia
To: Beijing, China
Statistics
• is the study of the
collection, organization,
analysis, interpretation and
presentation of information
or data
• all aspects of data including
the planning of data
collection in terms of the
design of surveys and
experiments are used
mathematical roots
• Two mathematical results of Bayes Theorem
are,
–THE LAW OF LARGE NUMBERS, and
–THE CENTRAL LIMIT THEOREM
law of large numbers (LLN)
• Is a theorem
• describes the result of
performing the same
experiment a large
number of times
• thus the average of results
obtained should be close
to the expected value, and
• The result will tend to
become closer to reality
central limit theorem (CLT)
• the arithmetic mean of a
sufficiently large number of
iterates of independent
random variables, each with a
well-defined expected value
and well-defined variance,
will be approximately
normally distributed
• This will probably lead to a
conclusion of the reality
normal distribution
• suppose
– sample contains a large number of
observations,
– each observation being randomly
generated,
– in a way that does not depend on the
values of the other observations
– that the arithmetic average of the
observed values is computed,
– and if this procedure is performed
many times,
– then central limit theorem says the
computed values of the average will be
a “normal distribution“
– The normal distribution is commonly
known as a "bell curve"
Bayes' theorem
is,
• a result that is of importance
in the mathematical
manipulation of conditional
probabilities
• It is a result that derives from
the more basic axioms of
probability
• It is a result that can lead to
realistic conclusions
Bayesian (or epistemological)
interpretation
• probability measures a degree of belief
• Bayes' theorem then links the degree of
belief
• in a proposition before and after
accounting for evidence.
• example, take a proposal that a biased
coin is twice as likely to land heads than tails.
– Degree of belief in this might initially be
50%.
– The coin is then flipped a number of
times to collect evidence.
– Belief may rise to 70% if the evidence
supports the proposition
axioms of probability
probability is,
• a game of chance
• measures the real, physical
tendency of something to occur
• Measures how strongly one
believes it will occur
• draws on both these elements?
• interprets the probability values
of probability theory to answer
the above questions
real life probability
Event Or Occurrence Non-Random event Random Event
Weather Seasonal Change, Local
Climate
Precipitation, Temperature on
Specific Days
Car Accidents Safe or Unsafe Driving
Practices
Specific Cars or Conditions
Met on the Road
Customers at Mall Hours Open, Time of the Day Specific Pattern of Customer
Arrival
State Lottery Decision About Games
Available, Prices and Awards
Numbers Drawn or Winning
Patterns on Tickets
Plane Crash Ability of Flight Captains,
Security Measures
Plane Engine Failure,
Terrorists Intrusion
Student Grades Amount of Time of study,
Revision and Class
Preparation
Appearance of Specific
questions on Tests
the
searc
h
why MH370 flight missing?
• Its loss is a mystery
• Many theories, probabilities
and interpretations still
remain
• Inconclusive, inadequate
and intriguing
• Probably Bayes Theoretical
approach would help
• Scientists tend to believe in
this theory as a probable
lead to discovery of the
missing plane
the theory that would not die
• “It‟s a very short,
simple equation that
says you can start out
with hypothesis about
something - and it
doesn‟t matter how
good the hypothesis
is,”
• - Sharon Bertsch McGrayne
how
because of,
• this character of the
formula
• forcing researchers to
change their hypothesis
with each new information
• that the probability
becomes more accurate
the proof
• Bayes’ Theorem “allows the
organisation of available data with
associated uncertainties and
computation of the PDF
• Scientists applied Bayes’ Theorem in
the Air France incident of 2009
• French scientist Pierre Simon Laplace,
helped locate German U-boats during
World War II and spot Soviet
submarines during the Cold War
• More recently, it is used in Google's
"driverless cars" project and in stock
market predictions
• PDF or the probability distribution
function, identified the target
location given these data"
Why Not?
• Advocates of this theory said it
was also used in the search for the
black box of the ill-fated Air France
flight 447, which crashed in the
vast Atlantic Ocean in June 2009
• What took two years for other
experts in the search for the black
box, took only five days for
consultants who applied the Bayes’
Theorem, to finally find the device
12,000 feet under water.
• In the current search for flight
MH370, it is “highly unlikely” that
Bayes’ Theorem is being applied.
Air France Accident summary
Date 1 June 2009
Summary Entered high altitude stall,
impacted ocean
Site Atlantic Ocean
near waypoint TASIL
[1]
3°03′57″N 30°33′42″WCoordina
tes: 3°03′57″N 30°33′42″W
Passengers 216
Crew 12
Fatalities 228 (all)
Survivors 0
Aircraft type Airbus A330-203
Operator Air France
Registration F-GZCP
Flight origin Rio de Janeiro–Galeão Airport
Destination Paris-Charles de Gaulle Airport
Inconclusive Theories about MH 370
• Sudden shut down of all
flight systems
• Sabotage by suspects
• Hijackers on board
• ‘Undesirable clients’ in the
cockpit
• Deep dive into ocean bed
• Disgruntled or insane flight
officer
• Ill equipped management
and security systems
• Etc., etc., etc…..
learn from experience
• Bayes' Theorem,
–is all about learning
from experience
–one would need
"reasonably accurate
past experiences" for
the theorem to work
–in other words, to
calculate accurately to
locate the plane
find the plane
• „go find the
plane, with
science in your
bag, enough of
spiritual and
emotional beliefs
and traditional
methods‟-masi
thanks
• Malaysian government reports/statements
• Wikipedia.com
• Malaysiakini Radio/TV and the Media
• Photo Gallery
• Gayesian experts like Sharon Bertsch McGrayne

Contenu connexe

Similaire à MH370 Search Using Bayes' Theorem

Illustrating uncertainty in extrapolating evidence for cost-effectiveness mod...
Illustrating uncertainty in extrapolating evidence for cost-effectiveness mod...Illustrating uncertainty in extrapolating evidence for cost-effectiveness mod...
Illustrating uncertainty in extrapolating evidence for cost-effectiveness mod...cheweb1
 
CS5032 Lecture 10: Learning from failure 2
CS5032 Lecture 10: Learning from failure 2CS5032 Lecture 10: Learning from failure 2
CS5032 Lecture 10: Learning from failure 2John Rooksby
 
Chapters 14 and 15 presentation
Chapters 14 and 15 presentationChapters 14 and 15 presentation
Chapters 14 and 15 presentationWilliam Perkins
 
Implied and Local Volatility for South African Derivatives
Implied and Local Volatility for South African DerivativesImplied and Local Volatility for South African Derivatives
Implied and Local Volatility for South African DerivativesAntonie Kotzé
 
Five Things I Learned While Building Anomaly Detection Tools - Toufic Boubez ...
Five Things I Learned While Building Anomaly Detection Tools - Toufic Boubez ...Five Things I Learned While Building Anomaly Detection Tools - Toufic Boubez ...
Five Things I Learned While Building Anomaly Detection Tools - Toufic Boubez ...tboubez
 
Risk And Uncertainty Analysis: A Primer for Floodplain Managers
Risk And Uncertainty Analysis:  A Primer for Floodplain ManagersRisk And Uncertainty Analysis:  A Primer for Floodplain Managers
Risk And Uncertainty Analysis: A Primer for Floodplain ManagersMichael DePue
 
Simple math for anomaly detection toufic boubez - metafor software - monito...
Simple math for anomaly detection   toufic boubez - metafor software - monito...Simple math for anomaly detection   toufic boubez - metafor software - monito...
Simple math for anomaly detection toufic boubez - metafor software - monito...tboubez
 
Introduction to data science
Introduction to data scienceIntroduction to data science
Introduction to data scienceSubrata Saharia
 
STSTISTICS AND PROBABILITY THEORY .pptx
STSTISTICS AND PROBABILITY THEORY  .pptxSTSTISTICS AND PROBABILITY THEORY  .pptx
STSTISTICS AND PROBABILITY THEORY .pptxVenuKumar65
 
Know Your Data: The stats behind your alerts
Know Your Data: The stats behind your alertsKnow Your Data: The stats behind your alerts
Know Your Data: The stats behind your alertsAll Things Open
 
Unit 6 input modeling
Unit 6 input modeling Unit 6 input modeling
Unit 6 input modeling raksharao
 
Gareth Digby: Systems-Based Approach to Cyber Investigations
Gareth Digby: Systems-Based Approach to Cyber Investigations Gareth Digby: Systems-Based Approach to Cyber Investigations
Gareth Digby: Systems-Based Approach to Cyber Investigations EnergyTech2015
 

Similaire à MH370 Search Using Bayes' Theorem (20)

Illustrating uncertainty in extrapolating evidence for cost-effectiveness mod...
Illustrating uncertainty in extrapolating evidence for cost-effectiveness mod...Illustrating uncertainty in extrapolating evidence for cost-effectiveness mod...
Illustrating uncertainty in extrapolating evidence for cost-effectiveness mod...
 
CS5032 Lecture 10: Learning from failure 2
CS5032 Lecture 10: Learning from failure 2CS5032 Lecture 10: Learning from failure 2
CS5032 Lecture 10: Learning from failure 2
 
Chapters 14 and 15 presentation
Chapters 14 and 15 presentationChapters 14 and 15 presentation
Chapters 14 and 15 presentation
 
Analysis
AnalysisAnalysis
Analysis
 
Implied and Local Volatility for South African Derivatives
Implied and Local Volatility for South African DerivativesImplied and Local Volatility for South African Derivatives
Implied and Local Volatility for South African Derivatives
 
Five Things I Learned While Building Anomaly Detection Tools - Toufic Boubez ...
Five Things I Learned While Building Anomaly Detection Tools - Toufic Boubez ...Five Things I Learned While Building Anomaly Detection Tools - Toufic Boubez ...
Five Things I Learned While Building Anomaly Detection Tools - Toufic Boubez ...
 
Mini datathon
Mini datathonMini datathon
Mini datathon
 
Ch9 slides
Ch9 slidesCh9 slides
Ch9 slides
 
Ch9_slides.ppt
Ch9_slides.pptCh9_slides.ppt
Ch9_slides.ppt
 
Risk And Uncertainty Analysis: A Primer for Floodplain Managers
Risk And Uncertainty Analysis:  A Primer for Floodplain ManagersRisk And Uncertainty Analysis:  A Primer for Floodplain Managers
Risk And Uncertainty Analysis: A Primer for Floodplain Managers
 
Simple math for anomaly detection toufic boubez - metafor software - monito...
Simple math for anomaly detection   toufic boubez - metafor software - monito...Simple math for anomaly detection   toufic boubez - metafor software - monito...
Simple math for anomaly detection toufic boubez - metafor software - monito...
 
Introduction to data science
Introduction to data scienceIntroduction to data science
Introduction to data science
 
STSTISTICS AND PROBABILITY THEORY .pptx
STSTISTICS AND PROBABILITY THEORY  .pptxSTSTISTICS AND PROBABILITY THEORY  .pptx
STSTISTICS AND PROBABILITY THEORY .pptx
 
Signal and noise
Signal and noiseSignal and noise
Signal and noise
 
Know Your Data: The stats behind your alerts
Know Your Data: The stats behind your alertsKnow Your Data: The stats behind your alerts
Know Your Data: The stats behind your alerts
 
Unit 6 input modeling
Unit 6 input modeling Unit 6 input modeling
Unit 6 input modeling
 
Introduction to Metocean: Quantifying the impact and effect of weather and se...
Introduction to Metocean: Quantifying the impact and effect of weather and se...Introduction to Metocean: Quantifying the impact and effect of weather and se...
Introduction to Metocean: Quantifying the impact and effect of weather and se...
 
PM 1 to 4.pdf
PM 1 to 4.pdfPM 1 to 4.pdf
PM 1 to 4.pdf
 
Gareth Digby: Systems-Based Approach to Cyber Investigations
Gareth Digby: Systems-Based Approach to Cyber Investigations Gareth Digby: Systems-Based Approach to Cyber Investigations
Gareth Digby: Systems-Based Approach to Cyber Investigations
 
Harmel - Monitoring to Support and Improve H/WQ Modeling
Harmel - Monitoring to Support and Improve H/WQ ModelingHarmel - Monitoring to Support and Improve H/WQ Modeling
Harmel - Monitoring to Support and Improve H/WQ Modeling
 

Plus de masilamani ramasamy (20)

Klia2 facts
Klia2 factsKlia2 facts
Klia2 facts
 
Leadership exercises 4
Leadership exercises 4Leadership exercises 4
Leadership exercises 4
 
Leadership
LeadershipLeadership
Leadership
 
Presentation on pmp exam
Presentation on pmp examPresentation on pmp exam
Presentation on pmp exam
 
The bayes theory [autosaved]
The bayes theory [autosaved]The bayes theory [autosaved]
The bayes theory [autosaved]
 
Essentialsof Project Management
Essentialsof Project ManagementEssentialsof Project Management
Essentialsof Project Management
 
Basics of sSpply Chain Management
Basics of sSpply Chain ManagementBasics of sSpply Chain Management
Basics of sSpply Chain Management
 
Nine Things-in-the-PMBoK
Nine Things-in-the-PMBoKNine Things-in-the-PMBoK
Nine Things-in-the-PMBoK
 
Black belt vs PMP Friend
Black belt vs PMP FriendBlack belt vs PMP Friend
Black belt vs PMP Friend
 
Business+Leaders-quotes worth cosidering
Business+Leaders-quotes worth cosideringBusiness+Leaders-quotes worth cosidering
Business+Leaders-quotes worth cosidering
 
The Hospitality Industry-the trends,
The Hospitality Industry-the trends, The Hospitality Industry-the trends,
The Hospitality Industry-the trends,
 
Developing Facilitation Skills
Developing Facilitation Skills Developing Facilitation Skills
Developing Facilitation Skills
 
Basic Facilitation Skills
Basic Facilitation SkillsBasic Facilitation Skills
Basic Facilitation Skills
 
Presentation1
Presentation1Presentation1
Presentation1
 
Leaders develop leaders
Leaders develop leadersLeaders develop leaders
Leaders develop leaders
 
First marketing workshopr of sridaya
First marketing workshopr of sridayaFirst marketing workshopr of sridaya
First marketing workshopr of sridaya
 
Bpm overview
Bpm overviewBpm overview
Bpm overview
 
Marketing management-ppt
Marketing management-pptMarketing management-ppt
Marketing management-ppt
 
Abilitities of leaders
Abilitities of leadersAbilitities of leaders
Abilitities of leaders
 
Successful project
Successful projectSuccessful project
Successful project
 

MH370 Search Using Bayes' Theorem

  • 1. the flight MH370 & Bayes theory Masilamani Ramasamy - a scientific management expert (misilamani@yahoo.com) Masilamani Ramasamy Consultant Trainer
  • 3.
  • 4. Bayes' Theorem • also called, Bayes' law or Bayes' rule • based on probability theory, mathematics and Statistics • named after Thomas Bayes (/ˈbeɪz/ 1701–1761) • Bayes suggested the theorem to update beliefs • Significantly edited and updated by Richard Price • Bayes' theorem “is to the theory of probability what Pythagoras's theorem is to geometry” -Harold Jeffreys
  • 5. probability theory rooted in, • the science of Statistics • the science of Mathematics • the probability of an event occurring • the analysis of random phenomena
  • 6. statistical theory gives, –ways of comparing statistical procedures –a best possible procedure for statistical problems –guidance on the choice between alternative procedures Passenger Identities Number China/Taiwan 154 Malaysia 38 India 5 Indonesia 7 Australia 6 France 4 USA 3 New Zealand 2 Canada 2 Russia 1 Italy 1 Netherlands 1 Austria 1
  • 7. statistical theory covers – Statistical decision methods and problems – Statistical inferences or conclusions – actions and deductions To satisfy the basic principles stated for these different approaches. STATISTICS(FACTS) Aircraft: Malaysian Air Lines Aircraft Age: 11years 10 months Aircraft Type: Boeing 777 ER Flight No: MH370(shared with China Southern Airlines) Passengers travelled: 227 Crew Travelled: 12 Nationalities: 14 Take Off Time: 12.41 am From: Kuala Lumpur, Malaysia To: Beijing, China
  • 8. Statistics • is the study of the collection, organization, analysis, interpretation and presentation of information or data • all aspects of data including the planning of data collection in terms of the design of surveys and experiments are used
  • 9. mathematical roots • Two mathematical results of Bayes Theorem are, –THE LAW OF LARGE NUMBERS, and –THE CENTRAL LIMIT THEOREM
  • 10. law of large numbers (LLN) • Is a theorem • describes the result of performing the same experiment a large number of times • thus the average of results obtained should be close to the expected value, and • The result will tend to become closer to reality
  • 11. central limit theorem (CLT) • the arithmetic mean of a sufficiently large number of iterates of independent random variables, each with a well-defined expected value and well-defined variance, will be approximately normally distributed • This will probably lead to a conclusion of the reality
  • 12. normal distribution • suppose – sample contains a large number of observations, – each observation being randomly generated, – in a way that does not depend on the values of the other observations – that the arithmetic average of the observed values is computed, – and if this procedure is performed many times, – then central limit theorem says the computed values of the average will be a “normal distribution“ – The normal distribution is commonly known as a "bell curve"
  • 13. Bayes' theorem is, • a result that is of importance in the mathematical manipulation of conditional probabilities • It is a result that derives from the more basic axioms of probability • It is a result that can lead to realistic conclusions
  • 14. Bayesian (or epistemological) interpretation • probability measures a degree of belief • Bayes' theorem then links the degree of belief • in a proposition before and after accounting for evidence. • example, take a proposal that a biased coin is twice as likely to land heads than tails. – Degree of belief in this might initially be 50%. – The coin is then flipped a number of times to collect evidence. – Belief may rise to 70% if the evidence supports the proposition
  • 15. axioms of probability probability is, • a game of chance • measures the real, physical tendency of something to occur • Measures how strongly one believes it will occur • draws on both these elements? • interprets the probability values of probability theory to answer the above questions
  • 16. real life probability Event Or Occurrence Non-Random event Random Event Weather Seasonal Change, Local Climate Precipitation, Temperature on Specific Days Car Accidents Safe or Unsafe Driving Practices Specific Cars or Conditions Met on the Road Customers at Mall Hours Open, Time of the Day Specific Pattern of Customer Arrival State Lottery Decision About Games Available, Prices and Awards Numbers Drawn or Winning Patterns on Tickets Plane Crash Ability of Flight Captains, Security Measures Plane Engine Failure, Terrorists Intrusion Student Grades Amount of Time of study, Revision and Class Preparation Appearance of Specific questions on Tests
  • 18. why MH370 flight missing? • Its loss is a mystery • Many theories, probabilities and interpretations still remain • Inconclusive, inadequate and intriguing • Probably Bayes Theoretical approach would help • Scientists tend to believe in this theory as a probable lead to discovery of the missing plane
  • 19. the theory that would not die • “It‟s a very short, simple equation that says you can start out with hypothesis about something - and it doesn‟t matter how good the hypothesis is,” • - Sharon Bertsch McGrayne
  • 20. how because of, • this character of the formula • forcing researchers to change their hypothesis with each new information • that the probability becomes more accurate
  • 21. the proof • Bayes’ Theorem “allows the organisation of available data with associated uncertainties and computation of the PDF • Scientists applied Bayes’ Theorem in the Air France incident of 2009 • French scientist Pierre Simon Laplace, helped locate German U-boats during World War II and spot Soviet submarines during the Cold War • More recently, it is used in Google's "driverless cars" project and in stock market predictions • PDF or the probability distribution function, identified the target location given these data"
  • 22. Why Not? • Advocates of this theory said it was also used in the search for the black box of the ill-fated Air France flight 447, which crashed in the vast Atlantic Ocean in June 2009 • What took two years for other experts in the search for the black box, took only five days for consultants who applied the Bayes’ Theorem, to finally find the device 12,000 feet under water. • In the current search for flight MH370, it is “highly unlikely” that Bayes’ Theorem is being applied. Air France Accident summary Date 1 June 2009 Summary Entered high altitude stall, impacted ocean Site Atlantic Ocean near waypoint TASIL [1] 3°03′57″N 30°33′42″WCoordina tes: 3°03′57″N 30°33′42″W Passengers 216 Crew 12 Fatalities 228 (all) Survivors 0 Aircraft type Airbus A330-203 Operator Air France Registration F-GZCP Flight origin Rio de Janeiro–Galeão Airport Destination Paris-Charles de Gaulle Airport
  • 23. Inconclusive Theories about MH 370 • Sudden shut down of all flight systems • Sabotage by suspects • Hijackers on board • ‘Undesirable clients’ in the cockpit • Deep dive into ocean bed • Disgruntled or insane flight officer • Ill equipped management and security systems • Etc., etc., etc…..
  • 24. learn from experience • Bayes' Theorem, –is all about learning from experience –one would need "reasonably accurate past experiences" for the theorem to work –in other words, to calculate accurately to locate the plane
  • 25. find the plane • „go find the plane, with science in your bag, enough of spiritual and emotional beliefs and traditional methods‟-masi
  • 26. thanks • Malaysian government reports/statements • Wikipedia.com • Malaysiakini Radio/TV and the Media • Photo Gallery • Gayesian experts like Sharon Bertsch McGrayne